There seems to be a massive demand for AI-generated voices. Companies are trying to sound more “human” than ever, and everyone has a different definition of that. When you try to program “humanness” into AI-generated voices, you can generate a lot of variety, which is important for both creative and enterprise use cases. Investors are ready to back startups in this area as there are a lot of opportunities for both research and market expansion. For companies, there is room to play around with accuracy, expression, and tone of these voices and master one or more categories.

In focus

Wispr Flow is rolling out a new model for quality improvement

Wispr Flow is one of the most popular transcription apps out there. However, the drawback of this popularity is that when something goes even slightly wrong, people start to take notice. Over the last few weeks, many people have complained about Wispr Flow’s accuracy and model quality dropping significantly.

In June, co-founder and CEO Tanay Kothari asked users to express their biggest gripes with the tool. One of the top requests was to improve accuracy.

In a chat with me last week, Kothari said that a year ago, the company decided that it couldn’t just put a band-aid on quality issues with quick fixes. The right way was to introduce a better model that solves current issues.

“We started realizing about a year ago that for these issues, the right answer is not to kind of put a Band-Aid and make those small changes. It comes back to why the model is making these kinds of mistakes, and realizing it hasn’t been trained with more intelligence. To solve this problem, we needed to train a bigger model,” Kothari said.

He said that the new model, which will likely improve quality significantly, is already rolled out to some users, and in the coming weeks will be available to all. The company also plans to better communicate what model changes can and can’t do for users and why there are quality drop-offs for some users and use cases.

New research lab and company direction

Wispr Flow is also taking steps not to be just a transcription company. It hired Ariya Rastrow, a researcher who worked on Alexa, as chief scientist to establish Wispr Advanced Interfaces Labs. Under this project, the company will research new models and new modalities of interaction with AI.

Kothari said that he and his co-founder Sahaj Garg have been thinking about cutting-edge research and new interfaces for a long time. But the challenge was to hire and assemble a team of people capable of working on these problems.

The new lab team comprises 13 people now and will grow bigger. The Wispr CEO said that the company is able to attract top talent as researchers want to work on cutting-edge product problems.

The aim with this lab is to build a new generation of models with reasoning models or intelligence behind them to serve new interaction paradigms.

“I think the whole generations of models that we're going to build are primarily going to be in the human-to-computer interaction space. Right now, it's just voice in, text out, and then you want to add a lot more modalities on both sides. We will use our own architecture and data for that,” Kothari said.

The company is not immediately thinking of building an API business out of its own models. However, it plans to publish more about research findings in the coming months.

As I reported last week, the company is also set to introduce a meeting note-taking feature and take on the likes of Granola, Firlies and Read AI. With this launch, the company could be taking the direction of being an executive assistant.

However, getting into meeting note-taking means the company is set to collect more data from you. Kothari told me that Wispr wants to be respectful and clear about how much context users are ready to give to the app, and if need be, take a step back.

With these moves, Wispr Flow is going to look like a different company, competing in various domains rather than being just a dictation startup.

Experiments

I recently started testing the Vocci ring. My initial impression was that it might be designed for personal note-taking, just like Sandbar’s Stream or Pebble 01 Index ring. However, strangely, it is built more like a meeting notetaker, as you start and stop recording with a button.

The ring is good at capturing audio and transcribing conversations. But the software is not ideal for when you need to take short notes. Second, I like meeting notetakers for in-person conversations to be visible to the other person. With this ring, the recording light is barely visible to me. Plus, it is easy for other attendees not tof notice a ring if a user doesn’t inform them that they are recording the meeting.

While this could count as a feature for some, as we form our meeting etiquette in the AI world, people would be wary of devices like these.

Numbers game

• ElevenLabs reached the mark of $600 million ARR in June. This marks another quarter with the company progressing towards that elusive $1 billion ARR mark. The company’s time to reach the next $100 million milestone has constantly been reducing.

Quick Bytes

• A few weeks ago, I said OpenAI is going to think more about being a strong enterprise player in voice AI. The company released two transcription models in the week called: GPT-Live Transcribe for real-time transcription, and GPT-Transcribe for batch mode.

• RHEI debuted a new agentic platform called Made for Music, which is designed for marketing a new track or an album release. The platform can create the press release, playlist pitch, cover art, and short-form promo assets.

• Friend released a new version of its pendant, which comes with a speaker and a pre-loaded voice that you can’t change. The company told Wired it is using OpenAI’s model to power the companion. We could see more gadgets in similar lines.

• A report by 404 Media highlighted that since Spotify’s AI music labeling is not working as intended, alternative sources such as SoullessMusic.com and Sloptracker have cropped up for users to check for AI-generated tracks. Increasingly, users are demanding more accountability from streaming platforms.

• There is a new class action lawsuit against Granola alleging the app is secretly recording meetings without users’ consent. Granola typically asks users to inform other parties that they are transcribing the meeting. The case highlights the lack of tools that can do this on behalf of users.

• India’s Sarvam released two new models for voice AI: Bulbul v4 for speech generation with better vocal range and Saaras v4 for automatic speech recognition with support for multi-speaker diarization.

Deals corner

Fish Audio ($50 million): The company started as an individual project from an NVIDIA researcher to generate better voices. Largely working in the text-to-speech domain, it has 8 million users and $21 million in ARR

• Investors: Coreline Ventures and Capital Today (lead), 359 Capital, Parable, Play Time, Alphalist Partners, Bayhouse Ventures, Carya Venture Partners, and HF0.

Smallest AI ($13 million): A text-to-speech startup that closed its funding this week. Smallest concentrates on smaller models for specific languages and dialects.

• Investors: Seligman Ventures (lead), Sierra Ventures, and 3one4 Capital.

Encore AI ($30 million): The startup observes voice calls and other customer conversations to train voice agents that can work with call center executives.

• Investors: Team8 (lead), Planven, Lukatz, and Garage.

Thank you for tuning in. Keep listening.

This newsletter is by Ivan Mehta, a freelance reporter at TechCrunch. It covers AI and technology in voice, audio, and music. Email: voiceaiweek@gmail.com or im@ivanmehta.com

The company is set to introduce a meeting note-taker (likely a Granola-style one) in the near future, its changed terms suggest.

The text in the changed terms suggests that the meeting notetaker will transcribe meetings, summarize them, identify speakers, and create notes and action items based on that.

Here is what the changed terms say related to notetaking:

"Notetaker" means Wispr's AI-powered meeting transcription, note-taking, summarization, speaker identification, action item generation, meeting insights, and related meeting assistance features made available through the Services.

The startup is currently testing the tool internally, but there isn’t a fixed timeline for the launch.

There isn’t a dearth of notetakers on the market. There are Granola, Read AI, Fireflies, Fathom, Otter, and Quill. Then there are tons of local, one-time purchase, and open-source notetakers.

With its meeting notetaker, Wispr gets to tell users that instead of paying for transcription and notetaking separately, you can just get one product. Though sometimes, people like an application more if it does one thing well.

Notetaking might be a crucial step for Wispr Flow to be a voice assistant or a full-stack voice company. The startup’s transcription product already learns the way you write emails and messages. With added context from meetings, it can easily create follow-up drafts that are a close resemblance to what you might have written.

Meeting notetaking is a table stakes feature for most productivity apps. It is not hard to plug in a speech-to-text model, transcribe a meeting, and summarize it using LLMs. The North Star for these apps is executives taking action and making decisions based on action items and insights from meeting notetaking. That is a more valuable and sticky asset for startups in the space.

Wispr Flow is valued at $700 million as per the last round. Bloomberg reported that the company is raising a new round at a $2 billion valuation. At that valuation, Wispr would want to be more than just a dictation company.

The startup’s founder, Tanay Kothari, has talked about chasing the vision of Jarvis, Tony Stark’s fictional assistant. That assistant needs to know more about users to execute commands when they talk to it. Context of what users do outside emails and in conversations is important. Meetings are the first step of that ladder to capture additional context for Wispr.

Signal

There are hundreds of hardware and software tools that are ready to “hear” what you say, but they are not yet fully ready to “do” what they heard from you. Voice updates from OpenAI and Anthropic, coupled with Genspark’s meeting notetaker hardware, suggest that companies are busy trying to acquire customers at a voice capture layer. Most of these companies claim users “get work done” through their voice layer, but integration with other software and commands results in final tasks not being smooth yet.

In focus

OpenAI and Anthropic update their voice stack with an agentic slant

Silicon Valley has been chasing the vision of an ever-present and efficient assistant like Jarvis in the Iron Man movies. The assistant listens to you perfectly and executes tasks. We do not have power suit-fetching voice assistants yet, but they can definitely fetch information and reschedule meetings.

Last week, Anthropic updated Claude’s voice mode, which now allows users to choose from Opus, Sonnet, and Haiku. This means that the company updated the intelligence aspect of Claude. Notably, there is no change in the voice model itself, so if you had complaints about how it talks to you, there won’t be any difference.

Another big difference is that the voice assistant can access app integrations like Gmail, Calendar, Notion, and Canva to execute tasks. It can rearrange your meeting or create a design document.

OpenAI, which had updated its conversational chops a few weeks ago, decided to update the ChatGPT desktop app, so the voice assistant can access Codex and Work to trigger agents and have them work in the background without you having to stare at the screen.

The promise of these updates is big, as it allows you to complete work without typing anything or staring at the screen. A startup founder said that he talked to Codex remote on mobile while he was on a hike.

There are a few hindrances, though; you will need to check on agents periodically to see if they are working as intended. For Codex, you will need to remember thread names, as voice runs in a different thread and needs you to call out other agents explicitly.

I partially blame this on OpenAI’s very confusing ChatGPT desktop app, where you don’t know the difference between Codex, Chat, and Work unless you explicitly remember what task you want to complete in what section.

As compared to OpenAI, Anthropic’s voice agent is useful for more general tasks, but there are some teething and latency issues as people switch between models.

AI companies are finally getting their voice features out of the conversation mode, which was useful for brainstorming and saving a transcript. However, with the new features, users can at least begin to do some tasks reliably.

My experiments

I have had brainstorming chats with ChatGPT’s new voice mode on my phone, but I wanted to put it through a different test of longer conversation and memory retention. My gym usually gets crowded during the evening, and I missed my usual time window because of work. I didn’t want to go to the gym, so I asked ChatGPT voice mode to prepare a mixed upper body and cardio workout.

Some exercises were repetitive, and I asked ChatGPT to swap them out. After a bit of back-and-forth, we settled on a routine.

The next task for ChatGPT was to stay with me during the workout and guide me. The first pass I tried was asking it to set an internal timer of 30 seconds per exercise. This went downhill fast. The timer failed multiple times, and ChatGPT told me that 30 seconds were over within 10-15 seconds of starting an exercise block. I pointed out the mistake a few times, and after that, ChatGPT said that it couldn’t do timers.

The alternative I found was that I would run my own timer. Then, once an exercise block was finished, I would ask ChatGPT to tell me the next exercise. This was good during the upper body block. However, after that, ChatGPT moved straight to cooldown rather than cardio, and I had to remind it of that.

The voice mode is good for conversation, but it needs to have a better memory and agentic tools to help with tasks on the phone. I’ll test this once again in a few weeks.

Feature that everyone should use

I recently started re-testing Willow’s dictation app on desktop. One neat feature in the latest update is that when you edit a particular word repeatedly, the app can detect the change and will ask you to add that to the dictionary.

Numbers game

Deezer said that for the first time AI-generated music crossed 50% of daily uploads on the platform. The company noted the volume of songs has gone up to over 90,000 AI tracks every day. Here is how the volume of AI songs has increased on Deezer”

Quick Bytes

• Suno’s troubles continued last week as TechCrunch reported that a November breach contained records of 55.3 million users, according to the Have I Been Pwned website. Separately, thousands of musicians joined a class action lawsuit against Suno and Udio.
  

• Amazon aims to switch models powering Alexa+ conversations as AWS costs could go up to $1.7 billion this year, according to Business Insider. Anthropic models are just not sustainable for Amazon, and it could use its own models.

• LG and Ericsson announced a partnership to bring voice AI directly to the network layer. We have seen Indian telco Reliance Jio make such announcements too. While the tech sounds interesting, we need to see what privacy challenges it can bring.

• Most Meeting notetakers focus on post-conversation knowledge and tips based on the transcript. Now, Otter wants to coach people during the conversation with its Live Assist feature. In the past, we have seen the likes of Cluely use a similar interface for real-time coaching.

• At a $700M valuation, you can’t just be a dictation company. Wispr Flow announced a new Interfaces Lab led by new chief science officer Ariya Rastrow, who worked on Amazon Alexa. He wrote that Wispr will focus on how users can interact with AI better, starting with the voice modality.

Deal corner

• Cast Insights ($4.5 million): A platform that provides audio insights from television, radio, podcasts, and livestreams for newsrooms and consultancies

  • Investors: Abstract Ventures (lead), SRI Ventures (lead), LocalGlobe, Martin Audio, Syndicate (VC)

Partner Spotlight with Atomik Growth

In a chat with a16z, ElevenLabs CEO Mati Staniszewski talks about how he and his co-founder, Piotr Dabkowski, got inspired to solve for badly dubbed movies and created a company that thinks that voice is the new interface for human-computer interaction.

Sponsored content

Thank you for tuning in. Keep listening.

This newsletter is by Ivan Mehta, a freelance reporter at TechCrunch. It covers AI and technology in voice, audio, and music. Email: voiceaiweek@gmail.com or im@ivanmehta.com

Into OpenAI’s hardware play with voice at center

We have seen many rumors about OpenAI’s hardware play with a pair of earbuds and a phone reportedly in the works. Last week, Bloomberg reported that the company’s first device could be a smart portable speaker with a camera that acts like an AI companion.

The speaker will have different sensors to understand your environment, along with a camera. Just like many AI assistants, OpenAI’s idea sounds like a ploy to collect as much context as possible about you and your surroundings in order to personalize its outputs.

More than 900 million people use ChatGPT for work, productivity, travel planning, shopping, and more. OpenAI would hope that the speaker can carry that context over and assist users in all kinds of tasks.

Here is what Bloomberg said about how the device would work:

OpenAI envisions the device anticipating needs, surfacing information proactively and serving as an expert on its user, they said. Though the speaker is designed to stay in the home, it will be easy to move around the house.

The weird bit Bloomberg reported is that the speaker will have mechanical components that move while responding, so users feel that it is alive. That description makes the device sound creepy rather than helpful. I would want my smart speaker to be efficient rather than emotive.

The speaker will directly go up against the likes of Amazon Echo, Sonos, and Google Nest, which have been around for a long time. OpenAI’s bet would be that with more sensors and an assistant that knows more about you, the speaker would attract many users.

The company has already deployed a new voice model, called GPT-Live, which sounds more expressive and human-like. The model is also better at holding conversations with improved turn-taking and interruption handling. The speaker can use a model from that family.

We have heard “AI is the new interface paradigm” many times. For OpenAI, voice might be the modality that unlocks it. The company is working with Jony Ive’s LoveFrom, hoping that the former Apple design head’s experience of creating an iconic new category of devices can wave his magic wand once again.

There are a few hurdles, though. On paper, the ChatGPT app on your phone can do many of the tasks a speaker can do, and there is no clear advantage to having an OpenAI speaker. Most smart speakers don’t have a camera, and in a world where users are wary of AI’s privacy implications, they might not want to have a device that can see what you do all the time from a company that has been accused of privacy infringement.

Smart speakers are good at a limited number of tasks, like setting timers or alarms, adding reminders, or playing songs. They can also control your smart home devices, but that part is unreliable at times. That is because either the speaker doesn’t understand what you said, or there is a breakdown in many pieces of software that connect your bulb to your speaker. AI or no AI, the basic expectation from any smart speaker would be that it can do all of the above without failing.

A more recent hindrance is Apple’s case against OpenAI, alleging that former employees now working at the AI giant stole trade secrets. The court procedure could take a long time, and it might not prevent OpenAI from releasing devices just yet, but there could be future roadblocks, thanks to the case.

AI devices haven’t really taken off in the mainstream yet. OpenAI would hope that it could breach this barrier.

Signals and Experiments

Listeners prefer multi-character AI voices over a human narration: Survey

Spoken conducted a survey of over 1,000 listeners in the U.S. about audiobooks narrated by its AI. The company said that 61% of people mistook AI narration for human narration. Spoken has a Multi-Cast tech with which it can power various characters for book enactments. The company pitted the tech against a single-person human narration.

Key observations from this survey:

• The AI narration had a higher percentage favorability at 61% vs 53% for human narration.

• 31% of listeners said that they would give AI narration a go. The number jumped to 65% after they heard a sample.

• 46% of people who heard the AI version showed an intent to purchase a title, but human narration rated higher, with 49%.

The audio publishing industry reached $2.43 billion in 2025. For smaller publishers or authors who might not have a budget for voice actors, AI could be helpful.

Platforms are offering tools for users to make creation easier. Apple started a digital narration program in 2023. Last year, Audible partnered with publishers by offering over 100 virtual voices to increase its catalog with rapid audiobook creation.

ElevenLabs offers a product for audiobook creation, along with incentives for distribution on its platform. Spotify started accepting books created on ElevenLabs in 2025 and launched a creation platform powered by the voice AI company’s model this year.

Book creation platform Inkfluence AI noted that audiobook creation increased 5x from March to May 2026. As the speed of creation and the variety of voices increase, many authors and publishers might not opt for human narration, impacting voice actor jobs.

Quick Bytes

• Another day, another Suno controversy. 404 Media reported that a hack revealed the company scraped services like YouTube, Deezer, and lyrics site Genius to train its models. Suno previously admitted that it used public sources to get data for its model, but maintains that it is within fair use rights.

• Avatar company Synthesia launched the latest version of its dubbing product, which lets teams refine transcripts and perform edits without spending tokens. The company said that its improved lip-syncing is better at syncing small mouth movements.

• Dictation company Willow made its transcription plan unlimited for base use. The company launched two new models, Willow Frontier Mini and Willow Frontier Pro. The mini model will power the free tier, and the Pro model is for faster and more accurate transcription for paid tiers.

• Local transcription is really useful when your internet is choppy. Communication coaching app Hedy said that it has improved its local dictation abilities on both iOS and macOS.

• Digital audio workstation (DAW) company BandLab acquired an AI music creation studio, Aiode, to expand its licensed audio-to-audio offerings. Aiode claims that its models are created in collaboration with musicians.

Deal Corner

• Rime ($24 million): The voice AI model company raised a Series A

  • Investors: M13 Ventures (lead), Twilio Ventures, Corazon Capital, Unusual Ventures

• Overtone ($18 million): Hinge founder’s new voice-based dating app raised a seed round

  • Investors: FirstMark Capital, Pace Capital, and Match Group.

Partner Spotlight with Atomik Growth

In a chat with a16z, ElevenLabs CEO Mati Staniszewski talks about how he and his co-founder, Piotr Dabkowski, got inspired to solve for badly dubbed movies and created a company that thinks that voice is the new interface for human-computer interaction.

Sponsored content

Thank you for tuning in. Keep listening.

This newsletter is by Ivan Mehta, a freelance reporter at TechCrunch. It covers AI and technology in voice, audio, and music. Email: voiceaiweek@gmail.com or im@ivanmehta.com

Before the deal, the company focused on building emotionally intelligent voice and language models and a pipeline for emotional labeling and metadata from audio. After the deal, under new CEO Andrew Ettinger’s leadership, the company has taken a new direction.

In a chat with me, Ettinger said the company now positions itself as a neutral player by becoming an emotionally intelligent voice infrastructure company.

Hume developed models around semantic space theory, measuring 48 emotions. While the models and pipeline were robust, the company realized that there was high value in its valuation and the data processing framework infrastructure.

“Our infrastructure was something that no one else in the world had because of the level at which we measured and got to that granularity. And so we started licensing that infrastructure out, and have built a fantastic business on top of that in the last five or six months,” Ettinger said.

“You could look at us as Switzerland, where we’re trying to provide infrastructure. The fact that we built this infrastructure with our research for our own models gives us a really unique perspective from first principles to explain and show people how we did this across hundreds of millions of hours of data and human judgments and emotion collections along with an evaluation infrastructure,” he added.

Leaning more into being a model-agnostic infrastructure company, Hume AI is releasing a new benchmark, called Real World VoiceEQ, to evaluate models on emotional intelligence and how “human” they are. The company tested more than 40 models for 60 metrics like Automatic Speech Recognition (ASR), Text-to-Speech (TTS), Speech-to-Speech (S2S), and Speech Understanding. It grounded its results in over a million evaluations, made by 10,000 participants, across the world using its Kairos platform, which is in private preview.

Here are some of the top observations from the study:

• Models can specialize in one ability, like technical accuracy, and falter in another, like emotional intelligence. This means we will see more specialized models. Plus, one-size-fits-all benchmarks are not enough to evaluate different features of a model.

• Voice models have become better at speaking and understanding the emotions of other speakers. However, they still don’t understand nuances like tonality and pacing.

• As we have covered in my story about Treble x Hugging Face benchmarks to measure model performance for real-life scenarios, evaluation metrics often ignore situations like noisy environments, overlapping speakers, and emotional expression.

• One of the biggest challenges for the voice AI industry is speech-to-speech models that listen and speak at the same time.

The study also revealed that even with accent variation in English, models can struggle to understand speech. Here is a passage from the research paper that suggests the error rate goes up 1.5x–2x for South Asian languages:

The accent results show that the native versus non-native gap is the most informative fairness axis. Across selected systems, the native-to-non-native WER gap ranges from +2.15 percentage points for ElevenLabs Scribe to +8.04 percentage points for AssemblyAI Universal-3 Pro. This gap is larger and more diagnostic than a standard-American versus other-regional-accents comparison, because the latter mixes easier native accents with harder non-native speech and hides the underlying fairness issue. Under Kachru’s Three Circles taxonomy, the second-language (Outer Circle) group — Indian, South Asian, Nigerian, Filipino, and related varieties — is the hardest condition for many models. In this subgroup, the top ten systems fall between 6.58% (Scribe) and 12.96% (Modulate Velma-2) WER, while the same systems score between 4.38% and 6.96% on native English.

Hume’s Olya Ossipova said that while studying benchmarks, the company observed that the current way of measuring voice AI models is often just to look at a transcript.

“When you think about expressiveness, what we found is that the model that performs best on expressiveness only ranks fourth in terms of taking action on that expressiveness. So there was actually like a very interesting gap that even if a model understands emotion, it doesn't really know what to do with it,” Ossipova said.

She also mentioned that there is an issue in real-life testing of models where LLM judges use objectivity rather than subjectivity, so you need to perform human evaluations as well. Ossipova said that peering into these problems will give model builders a way to make voice AI models more efficient.

Hume is currently running Kairos, which lets you automate evaluation models for this index, in a private preview. The company is also planning a new partnership with Hugging Face for benchmarks. There is a rising trend of evaluating voice models in different real-life situations, and also assessing models based on their emotional understanding.

OpenAI releases new conversational models to boost voice usage

The headline number from OpenAI this week was that more than 150 million people use ChatGPT’s voice features — voice mode and dictation. Assuming over 900 million people use ChatGPT weekly, that means roughly 16.7% are engaging with its voice features, giving us a clear sense of the massive consumer demand for audio-first tools. This gives us a rough idea of how large the consumer market for voice tools is right now.

OpenAI released a new pair of models last week, named GPT-Live-1 and GPT-Live-1-mini, to enhance ChatGPT’s voice mode for live conversations.

Architectural changes

The company’s naming of the voice model is a bit of a mess because the voice model was previously powered by a model named “Advanced voice model” (AVM). AVM was a pipeline of speech-to-text, a large language model, and text-to-speech. This meant that turn detection and interruption handling were clunky. Plus, you didn’t always get the answers that you wanted.

The company has adopted a new architecture to solve these problems. First, the GPT-Live series is a pair of full-duplex models. This means they can listen and talk at the same time without having to wait for someone’s turn. Second, reasoning, search, and agentic capabilities are separate to keep the conversation going and get access to the best knowledge models like GPT-5.5 for more accurate answers.

App usage

OpenAI has big ambitions with the model. The company said during a media briefing that with this kind of change to voice mode, it thinks of voice “as a kind of primary interface to computing.” With improved conversational skills, OpenAI is betting that its weekly active user base for voice will surge well past 150 million

While OpenAI didn’t provide any data about average session time, the team that built the feature said they had many conversations in voice mode spanning 30-40 minutes. With the new and improved voice mode, the company would like increased session times.

How will OpenAI use these models?

Apple, Google, and Amazon have tried to make their assistants sound more natural, conversational, and knowledgeable. While all these companies have their own devices, OpenAI has to rely on existing platforms for distribution. The new voice mode could set the stage for the company's rumoured devices like earbuds and a phone. Notably, OpenAI was sued by Apple this week over issues of trade theft and breach of contract. It might not derail the AI company’s hardware plans just yet, but we have to keep watching the case to know more.

What’s more, OpenAI plans to bring these models to its API for developers and enterprises. The call center and customer support market is hot. With models that can better handle interruptions, OpenAI could make itself a serious player in the market to compete with the likes of ElevenLabs, Deepgram, and Cartesia.

Signals and Experiments

I got access to the new ChatGPT Voice mode with the rollout and decided to put it to the test. The voice sounds more expressive, and you have the option to switch intelligence (Instant, High, and Medium) and languages directly from a quick settings menu present while using voice mode.

The model still occasionally talks over you, but when it realizes its mistake, it pauses much more naturally than the previous version. Overall, the conversation flowed much better

Earlier in the week, I had a conversation with ChatGPT voice mode about the impact of voice as a modality on app engagement, and the older model wasn’t able to fetch any data. The newer model cited some studies for me.

OpenAI claimed that you can get instant translations while speaking because of the full-duplex nature of the model. In practice, the latency works, but I wish I could say the same about the quality of voice and translation. I tried conversing with the new voice mode in English, and it was speaking Hindi; the pronunciation was that of a foreign-born Indian with a heavy accent, and the assistant spoke bookish Hindi at times that people don’t really use in real life.

The company said that the new voice mode supports “most spoken languages” but didn’t list them. I wish that multilingual coverage were better because I would like to use the assistant for translation when I travel.

The second issue for me was basic work that an assistant can do in voice mode. I asked it to list some points we discussed about newsletter topics, and it started narrating the points without actually listing them. A few times, it told me it could make a framework for graphics for certain stories, but it never ended up making one because it can’t do so.

OpenAI did show off the ability to give visual answers, but those are limited to sports schedules, maps, and weather. I wish the company would give the voice mode the ability to do more things, including creating documents. ChatGPT already has a way to transcribe your prompts, but it would be great if I could keep talking about tasks and the voice mode performs them in the background.

This week, OpenAI launched OpenAI Work, a new agent for project work, and also updated the ChatGPT desktop app to merge Work and Codex. With plugins and coding ability, I would really like to chat with the agent while I might be checking email or doing other tasks.

OpenAI has hinted that it wants people to use their voice for work-related tasks, and the new voice mode could just be the beginning of that path.

Quick Bytes

• Paris-based company Gradium extended its seed round of $70 million in December to another $30 million, bringing the total to $100 million. The company is focusing on building real-time speech models and solutions.

• Music industry bodies have come together to create a program for AI track labeling on streaming services with markers for “AI-assisted” and “AI-generated” tracks. Streaming services like Apple Music and Tidal have adopted their own labeling, while Spotify shows AI involvement in song credits.

• Speechify launched its Simba 3.2 model, which is currently topping the artificial analysis leaderboard. The model also has a competitive base pricing of $10 per million characters, which reduces to $6 per million characters at scale.

• Character.AI is thinking about audio drama with AI characters through its c.ai FM product. The company is currently testing these serialized dramas in a limited way, with professional writers penning stories and AI-generated characters narrating the script.

• Music companies are now trying ethical ways to train models. Music production platform Landr said that more than 30,000 artists have opted into its AI revenue share program. The company has also created a $1 million pool to fund advances for eligible tracks joining the program.

• Google Voice added transcription and note-taking capabilities that already exist within Google Meet. Users can press the “Notes” button during the call to get a summary, transcript, and action items.

Partner Spotlight with Atomik Growth

In a chat with a16z, ElevenLabs CEO Mati Staniszewski talks about how he and his co-founder, Piotr Dabkowski, got inspired to solve for badly dubbed movies and created a company that thinks that voice is the new interface for human-computer interaction.

Sponsored content

Thank you for tuning in. Keep listening.

This newsletter is by Ivan Mehta, a freelance reporter at TechCrunch. It covers AI and technology in voice, audio, and music. Email: voiceaiweek@gmail.com or im@ivanmehta.com

India’s ToneTag bets on its acoustic tech for local processing

After the short era of “Tokenmaxing”, companies using AI are being more cautious about their token budgets and trying to minimize costs. One way that has emerged in this frugal era is to offload part of the processing to local models and chips.

Even if companies can infer some queries without having to go to the cloud, they can start to save on costs. Indian fintech company ToneTag is doing this by processing acoustic signals and answering merchant queries about recent payments locally.

After the Unified Payments Interface (UPI) gained prominence as a digital payment medium in India, banks, payment providers, and fintech startups started deploying small speaker boxes with QR codes. These QR codes “announced” the payments merchants received, which was helpful for sellers who were new to digital payments or might not have a smartphone handy to check the status of a payment.

Using acoustic signals for AI processing

Until now, these boxes have been basic, with some functions like recalling recent payments and volume control. ToneTag, backed by Amazon, MasterCard, and Qualcomm, produces these boxes. The startup recently revealed a new project called eKosha, which uses AI to answer merchant queries.

The startup has a patent for using audio to process payments. At the core, the patent is about encoding and decoding data by embedding it into sound waves. While the original use of this was to facilitate payments between two devices, such as a feature phone and a payment device, the technology could be used to decode requests for language models.

ToneTag’s CEO, Kumar Abhishek, told me that phone processors are capable enough to convert all speech into digital signals like text and hand over the request to models behind Siri or Google Assistant. But for payment boxes, there needs to be a different strategy.

“When you start adding AI to IoT devices, where you have less compute, less power, and less memory, it's very difficult for you to keep listening to requests and process them. We were able to do this because we do analog filtering to make local processing possible,” Abhishek said in an interview.

What can this tech do?

With this implementation, merchants can ask anything about the box itself and troubleshoot problems. Plus, the AI assistant can answer questions about cash flow and recent transactions. This saves merchants from frantically checking bank apps to see the status of a payment or calling customer care to get basic information. The box supports 10 Indic languages and mixed-coded speech with plans for expansion.

ToneTag is also working with banks to put AI to work through cloud queries. With the device, merchants can ask banking-related queries to the device, and also get suggestions for financial products. With all these requests, the bank can gauge who might be interested in lending-related products.

“Because these boxes are conversational, banks could extend their MSME banking right on the box. A merchant can request a checkbook or a credit through voice commands. All tasks that needed a call to the relationship manager or a bank visit could be done directly from the box,” Abhishek said.

ToneTag’s new box has a design quirk for shopkeepers in a dedicated query button. This button serves two purposes: you don’t have to go across the room to start a query, ensuring queries are intentional rather than accidental activations or casual questions about the weather. Given that there is a mandatory button press for the query, merchants have to be close to the device (and, in turn, the mic), resulting in better quality of audio capture. The company said that the merchant's perception of the device and query volume saw an uptick after this implementation.

India has more than 23 million soundboxes deployed across merchants. The problem for fintech companies is merchants retaining soundboxes. Often, distributors offer a box from a company to merchants at a cheaper rate, but as subscription prices increase, merchants would move to another box because the switching cost is low. Abhishek and ToneTag, which aims to reach 500,000 merchants with the new eKosha box in the coming month, think that AI and embedded services will get merchants to stick around with the same box.

Quick Bytes

• Investors are bullish that meeting capture is one of the biggest unlocks of AI. After pouring millions into digital notetakers, they are looking at the physical space where the likes of Plaud and Viaim are competing. Now, another competitor, Pocket, which has sold over 130k devices, got $11 million in funding from Accel, Y Combinator, Peak XV, and ElevenLabs CEO and co-founder Mati Staniszewski.

• ElevenLabs, which has multiple celebrity voice partnerships, will power Gene Wilder’s voice for Netflix’s upcoming reality show “Wonka’s The Golden Ticket.” The company has gotten permission to use the voice from the Gene Wilder Estate, as per Variety.

• Suno is prepping to launch its API for music generation with licensed data. The end goal is to get blessings from the label for partnership and provide other toolmakers with the ability to create sounds or tracks. But legal trouble shows no signs of abating for the company as Winamp Group sued the AI company for unauthorized use of its data for model training.

• ElevenLabs’ value continues to soar. In a fresh tender offer, the voice AI company is looking at a $22 billion valuation, up from $11 billion during its fundraise in February.

• Some reports have suggested that a good chunk of AI-generated music is to game streaming service systems and generate revenue. Tidal is taking a stance against that by demonetizing AI music revenue. The company will tag fully AI-generated tracks, but it didn’t specify what kind of tech it might use to detect AI in songs.

Signals & Experiments

I wanted to escape Delhi’s heat, so I took a two-day trip to Dehradun. While I was there, I stayed in the Taj hotel, and the hospitality was generally great. On the first night, I asked the front desk to send me a toothbrush and a few water bottles. I didn’t get them for 40 minutes. I called them again. This time, I got water bottles, but no toothbrush.

First, I missed my robot friends from hotels in China. Second, can taking requests through voice or chat solve this? I have seen hotels ask you to contact them over WhatsApp through a QR code on the TV screen. But often they just have hotel room information in the chat.

If I can leave a voice note or type my request out, there is less of a chance of missing out on items in my request. This use case doesn’t particularly need large language models. Just a way to get my entry to the request queue to the front desk.

Partner Spotlight with Atomik Growth

In a chat with a16z, ElevenLabs CEO Mati Staniszewski talks about how he and his co-founder, Piotr Dabkowski, got inspired to solve for badly dubbed movies and created a company that thinks that voice is the new interface for human-computer interaction.

Sponsored content

Thank you for tuning in. Keep listening.

This newsletter is by Ivan Mehta, a freelance reporter at TechCrunch. It covers AI and technology in voice, audio, and music. Email: voiceaiweek@gmail.com or im@ivanmehta.com

Why Heidi is building specialized hardware for health

Health-focused voice AI companies have raised over $280 million in funding this year, including Assort Health’s recent $120 million Series C raise to scale voice agents that have handled over 190 million calls. Of that total, nearly $100 million has gone to companies that have a scribe product for notetaking and follow-up actions.

Hardware meeting notetakers are a rising category. Plaud, Anker, Vibe, and Pocket are focusing on professions like doctors and healthcare providers to rely on their gadgets to take notes or record patient sessions. These companies are also targeting all professions where in-person meetings and conversations are taking precedence.

In contrast, Australia-based Heidi Health is taking matters into its own hands by using its specialized models — trained to understand medical terminology better — and approaching clinics directly to deploy its scribe. The company also decided to build its own hardware scribe to improve transcription quality.

I spoke with the company’s CEO, Tom Kelly, about its approach, which has now scaled to handling 3 million weekly visits. Heidi began as a tool for a patient’s history and a chatbot built on top of that.

Building its own models

The company realized that healthcare providers spend a lot of work in admin-related tasks such as follow-ups and filling forms based on information patients provided. That is why Heidi started focusing on making its scribe top-notch. It also wanted to sell directly to clinics to get direct feedback and improve the product.

Kelly said that the company tuned models for medical use and claimed they are better than what frontier labs have to offer for speech-to-text and language summarization.

“In 2025, we struggled a lot with third parties having outages or variations in performance. So we would notice, for example, when a Frontier Lab launched a brand new model, they would be very excited about it. Everyone at Heidi would be disappointed because we would experience a degradation in the old model,” Kelly said.

He added that another problem was that Heidi didn’t have a full view of what happens to requests underneath. That is why the company started building and relying on its own fine-tuned models.

Heidi said that often, generalized models try to “guess” the medical term, which could be dangerous for patients. That is why the startup tries to ground its models in specialized medical terminology for different sectors.

Hardware play

Besides models, Heidi wants to own other parts of its voice stack, including hardware. Rather than relying on just the phone mic or third-party gadgets, the company built its own notetaking device for better transcription quality and, in turn, better follow-up actions.

“The most common cause of bad transcripts is having a bad microphone or a phone that’s put in a bad position. So it’s really important for us to never miss a doctor’s session. We want it to be high quality, and we want to be able to hear what they’re doing,” Kelly reasoned behind building its own device.

Heidi said that healthcare providers often move around a lot, and when the startup tested third-party mics, they weren’t accurate. When it started to test its own device, the results were much better, and it went on to develop its own hardware.

More than just a scribe

The company is also starting to test and deploy its own Wispr Flow-like dictation tool, but for clinical use. The idea is similar: users press a key on their desktop and transcribe what they need across apps.

The company’s aim is to run all of its processing offline in the future. Kelly said that it is possible to run reliable dictation or transcription on a device, but for running LLMs, you either need a dedicated device for a hospital with on-prem deployment or buy a machine with large RAM and storage. It also wants to tap into its origins and reason from a patient’s history to provide a correct diagnosis.

A Menlo Ventures report from 2025 suggested that AI note-taking in healthcare saw a $600 million spend. Investors are very interested in this category as clinics might not switch scribes easily, and startups that build a context graph for health over time will have a dominant market positioning.

Signals & Experiments

Amazon’s Alexa assistant has existed for nearly a decade in India. After its introduction in 2017 with support for English, it later added Hindi. Yes, it understands both languages and sometimes a mix of them. Alexa needs drastic improvements in understanding users.

I have a fan and a table lamp connected to the smart home ecosystem that I can control through Alexa. And often Alexa switches on one instead of another. And at times, it just plays a song instead of switching off the light. I barely ask Alexa for facts or news because it often takes me three to four times to just get my request right.

Last week, I reported for TechCrunch that Amazon is testing Alexa+ in Hindi in India. With this experiment, I hope Amazon is focusing more on accurately understanding Indian speech rather than just adding a “conversational” assistant.

Amazon and Alexa still command a large share of India’s smart home market. However, if you want an assistant to chat with for news, facts, music, and more, there are dozens of options out there with new ones launching every week. Thanks to newer AI models, some of them are good at understanding varied speech. Amazon will need to bring better models to create a stickier experience for consumers.

Quick Bytes

• Coval, a startup built by former Waymo engineer Brooke Hopkins, raised $28 million in a round led by Norwest. Voice AI is scaling up, and enterprises are in need of an infrastructure that tells them how agents are performing and if their money is being used effectively.

• Voice startup Speechify launched its dictation tools for iOS and Mac. With people building their own tools for dictation, the market is heating up. People in the tech industry might choose open-source or niche tools or create one for themselves. However, enterprise customers will tilt towards the likes of Wispr Flow, Willow, and Superwhisper for compliance reasons.

• Zoom is leaning into the ethos of agents working based on meeting notes and discussions with Zoommates. These new agents can create first-draft presentations, reports, project plans, proposals, and spreadsheets from conversations.

• Voice AI company Modulate released a new API to catch AI-generated music. There is a strong wave of AI-generated music making it to streaming services, and companies like Deezer have been fighting it hard with their own tools. Labs like ElevenLabs are adopting Google’s SynthID for generated audio watermarking. But it is still not a widely accepted standard across the board.

Off Topic

I love testing new apps, and I have been obsessed with this iOS weather app called Good Air. It is beautiful and has great transitions as you scroll to different times. Plus, it also tells me the AQI level near me. As someone who lives in one of the most polluted cities. That is important information, and at least I get to look at the horrible weather with a good interface.

Partner Spotlight with Atomik Growth

In a chat with a16z, ElevenLabs CEO Mati Staniszewski talks about how he and his co-founder, Piotr Dabkowski, got inspired to solve for badly dubbed movies and created a company that thinks that voice is the new interface for human-computer interaction.

Sponsored content

Thank you for tuning in. Keep listening.

This newsletter is by Ivan Mehta, a freelance reporter at TechCrunch. It covers AI and technology in voice, audio, and music. Email: voiceaiweek@gmail.com or im@ivanmehta.com

There are independent benchmarks like the Open ASR leaderboard, but they don’t always capture real-world scenarios with messy audio and noise. Plus, these kinds of leaderboards deal with clean audio in a near-field situation where a user is speaking into a phone or a headset.

This is why Hugging Face and Iceland-based voice simulation and testing company Treble decided to create a benchmark specifically for Far Field ASR (FFASR). It evaluates environments where a user is one meter or farther from the mic, such as conference room meetings or someone talking to a smart speaker.

Treble has a simulation where AI companies or device providers can test their models in real-life scenarios, such as various restaurants or cafes. The company told me that over the years, it has developed simulation tech that can create complex situations and test models much faster than before. Using these techniques, Treble can tell AI companies what kind of physics understanding they can build in voice models for better accuracy.

The startup’s CEO, Finnur Pind, said that until now, companies measured their models’ capabilities through the Open ASR leaderboard, which is built on clean audio. He also noted that it could be possible to prepare models for better benchmark performance, as the data is known.

“For Hugging Face and us, it was important to measure how ASR performs out in the real world in complex conditions with overlapping talkers, reverberation, and movement. We wanted to try to create a benchmark that tests these conditions,” Pind said.

The FFASR leaderboard has different datasets for comparison. There near-field speech sample of 8 hours as a base. Then, through Treble’s SDK, there are high, mid, and low signal-to-noise ratio datasets that test models in scenarios like the bathroom, living room, meeting rooms, and hallways with noises like coughing and the HVAC system. Other measurement parameters are of high, mid, and low signal-to-noise ratios with movement. All these tests are based on the Treble10 dataset that the Icelandic startup and Hugging Face collaborated on last year.

However, Pind mentioned that the core data is hidden to avoid gaming of the benchmark. He also mentioned that performance measurement on the old leaderboard had plateaued. And the new FFASR leaderboard will make way for a new kind of measurement.

When model companies test on this benchmark, they would get an average word error rate (WER) across datasets, and an RTFx score, which measures the duration of audio and the time it took for a model to process it. There is also a plot to show the correlation between the two.

“When you test out models on real-world conditions, the performance fluctuation is large for different scenarios. This new benchmark will show companies where their models have been lacking,” Pind said.

Looking ahead, Hugging Face and Treble plan to expand the platform. In a recent webinar, Hugging Face mentioned that in the future, it wants the community to contribute to benchmarking with new open-source datasets for far-field evaluation, ways to test hardware, new denoising algorithms, and different acoustic conditions to test out models. You can check out the benchmark here.

India’s telecom network Jio, owned by the Reliance conglomerate, comes up in every discussion around how it revolutionized data access and made it possible for many users to start using the internet en masse. The company, which is filing for an IPO, is now leaning heavily into voice to make AI accessible to its user base of over 500 million.

The company promises that people using Jio’s network will be able to access AI without needing any extra app and directly on calls. Users will be able to just call up the assistant by saying “Hey Jio” on the call and follow it up with a request to transcribe a call, create action items, place food orders, and more. This raises privacy questions about how Jio is listening in on the call for requests and how much data it retains. Will there be a way to turn off Jio AI for me?

Indian laws would allow users to transcribe a call for personal usage or safety, but usage beyond that without the consent of other people on the call can land you in legal trouble. Also, what happens when you place an international call in a territory where you need two-party consent for recording?

Jio showed a short demo of its voice AI functions during its Annual General Meeting (AGM) of shareholders, and I have questions about how some of that would work in real life.

The demo showed a person in the office on a call, asking Jio’s assistant to add “Rohan, Priya, and Kavya to this call,” and the assistant did its job. But how did it know which Rohan, Priya, and Kavya to add? Did it have access to the person’s phonebook? And if yes, are these the only contacts with these names? It seems unlikely because these are common first names in India.

Moments later, the person on a call with their family asks the Jio assistant to order a pizza. The company said that the assistant knew about the person’s usual outlet without specifying how. It even prompted the person to add Campa Cola, which is a Reliance-owned product. Will the assistant always promote company-owned products? What if I ask it never to recommend Campa Cola?

Jio’s premise is that users in India use voice a lot to chat with each other and with companies. Placing an AI assistant without having them install an app is one of the best ways to get distribution. There are clearly big questions about how the assistant would work, both in terms of execution and data privacy for users.

The company is also introducing voice AI across its apps and ecosystem for mobile recharge, customer support, installation help, roaming and eSIM activation. These use cases are app-based or call-based, so they seem much more plausible and consent-based than the on-network AI assistant.

Jio also showed off a new device called JioTeleFrame, which is a tablet that sits on a wall in your home and shows you reminders, schedules, and weather alerts. Jio calls them agents, but that’s a bit far-fetched. At one point, the demo showed two people talking about planning a festival lunch, and one person asked, “Hey Jio, did you get that?” and the tablet showed a potential lunch menu. Was the tablet listening all along? How would a guest know that a device was listening to them?

The company shows that the tablet can listen to you from a distance, which would be fantastic, because even my Echo devices with multiple mics make mistakes in understanding my voice at a few feet of distance.

The tablet can control the TV, smart home, place shopping orders on JioMart, and possibly talk to other Jio apps and devices. The idea of a central controlling and information device is not new, but in India, where a lot of information is fragmented, it will have to pull a lot of weight to work.

Jio has realized that voice could be a big part of India’s AI story. The company has resources and scale to reach millions of consumers. However, these products will need to perform in real life and not in demos, which would be a tough task to get to.

Signals & Experiments

AI companies are in a race to build AI voices that sound more “human.” These voices would ideally have more texture, timbre, expressiveness, and natural tonality. The core idea is that the more human these AI voices sound, the more likely that actual humans would want to talk to them for use cases ranging from customer support to therapy bots.

It is hard to quantify the “humanness” of a voice model. But that is what Vapi’s new experimental index aims to do. The voice AI pipeline company has built a site where users can listen to a pair of the same voice cloned to different models and pick one based on which one sounds more human. They can carry on doing this for different pairs. Based on these blind votes, Vapi is giving ratings along with a “humanness” score to these models.

At the time of writing this newsletter, xAI’s Grok TTS is the leader, followed by ElevenLabs’ Elevenv3 and Minimax’s Speech 2.5. But these ratings can change as more votes come in. The index also shows latency and pricing for comparison.

Quick Bytes

• There is a new bipartisan bill in the U.S. that would allow citizens to protect their voice from unlawful AI usage. Critically, Google and OpenAI are both supporting this bill. Notably, Denmark gave similar rights to its citizens last year.

• Voice AI company Bland raised $50 million in Series C funding led by Dell Technologies. Fortune noted that the company runs its own models on controlled infrastructure for regulated industries like healthcare. Bland now handles over 3.5 million calls per week and tackled over 175 million calls last year.

• Zoom is adding more voice integration to its suite. Users will be able to use voice commands to start meetings soon. The company also said that this year, it will roll out a feature that will intelligently tag speakers in a call.

• India’s Sarvam, which has multiple voice models and an agent platform, closed a $234 million funding round led by HCLTech. While the company is building large language models from the ground up, it has released multiple voice models that are proficient in Indian languages.

• Hardware voice AI notetaker Plaud said that the company is at $100 million ARR from its software subscription business. It has sold more than 2 million of its note-taking devices. Bloomberg noted that the company is targeting $500 million in sales, and new devices are on their way.

• Apple patched a bug in Beats Studio earbuds that could turn them into a tapping device, according to Malwarebytes. The report noted that attackers could listen in to coversations and also impersonate trusted headphones to other devices.

Partner Spotlight with Atomik Growth

In a chat with a16z, ElevenLabs CEO Mati Staniszewski talks about how he and his co-founder, Piotr Dabkowski, got inspired to solve for badly dubbed movies and created a company that thinks that voice is the new interface for human-computer interaction.

Sponsored content

Thank you for tuning in. Keep listening.

This newsletter is by Ivan Mehta, a freelance reporter at TechCrunch. It covers AI and technology in voice, audio, and music. Email: voiceaiweek@gmail.com or im@ivanmehta.com

Apple’s new Siri AI looks promising, but needs speech improvements

Apple announced the much-anticipated Siri AI (why not just Siri?) at its Worldwide Developer Conference (WWDC). The new Siri is a working iteration of what Apple promised two years ago — an assistant that can better understand you, get insights from the data on the device, look at the screen, and use that context to provide answers.

The company also said that users will be able to customize Siri’s voice by controlling pace and expressiveness. All voice companies and Assistant makers have been trying to get AI to sound more human. The demo of Siri Apple showed in the WWDC keynote video didn’t sound expressive, but this may change as the new betas roll out.

Another small but important announcement from Apple was an improvement in dictation with an on-device model. It’s not clear if the dictation features would equate to using an app like Wispr Flow, Willow, or Monolouge. However, as I have reported before, Apple broke how these apps worked with its iOS 26.4 upgrade. With its own dictation feature built into the default keyboard, users won’t have to switch apps to activate a dictation session.

There is also a dedicated app for the new assistant, which you can use as an alternative to ChatGPT or Gemini. But I haven’t got a chance to test it long enough.

After being on a weeklong waitlist, I got access to the new Siri just today. Here are my very early observations from a few hours of testing:

• Siri can understand me better, even with filler words and complex sentences. However, as soon as I finish one sentence, Siri fires up requests to fetch knowledge even when I am completing my request. When I said, “What are the World Cup matches for today, and where can I watch them?” Siri started searching for matches the moment I said “today.”

• For the above request, for some reason, Siri fetched matches between “Last Friday and Thursday.”

• The voice is not very expressive, but it is better than previous versions of Siri. Right now, the expressiveness dials are in preview, and there is no way to adjust them.

• I turned on the advanced dictation preview, but there is no specific indicator to tell me if I am using the new dictation or the old one.

I am going to keep testing the new Siri for the coming weeks and report my observations.

Model Behavior

Google releases a new live translation model with support for 70 languages

Google dropped a new live translation model powered by Gemini 3.5 with support for over 70 languages. In last week’s newsletter, I talked about using translation apps in China. The new model works on Google Translate, which might not work in countries like China, unless I download the models.

On the enterprise side, the company is debuting the model in Google Meet for enterprises next month. It noted that companies like Grab are experimenting with the new model for customer calls already. This will create a competition between Google and translation providers like Krisp, DeepL, and Palabra

Apple’s new on-device model for powering speech and dictation

Apple’s new Siri is powered by a series of models, including Apple Foundation Model (AFM) 3 and AFM 3 Core Advanced. It’s the second model that is powering the new expressive voice and dictation. The company said that the Core Advanced model is a 20-billion-parameter model, but it acts like a mixture of expert model with 1-4 billion active parameters.

Signals & Experiments

Robot’s slip of tongue

During my China tour, I visited a robotics company called Keenon. They had demoed various robots, and one of them was a humanoid with talking capabilities. When someone asked about a tech blog, it messed up the pronunciation and made it into profanity.

Tech companies might perfect all the robot movements, but numerous examples of chatbots going rogue have taught us that it is highly likely that robots might slip up while conversing with humans at any time. Companies will need to be extra careful as their robots will be out in the wild, and a customer who will experience this mishap might not be as forgiving as tech journalists looking at demos.

Quick Bytes

• Wispr Flow admits that the app’s performance hasn’t been up to date. The company said it is taking steps like scaling its infrastructure to tackle outages and rolling back changes to its autocorrect algorithm. It claimed to have bolstered internal tooling to catch issues quickly.

• India’s Equal AI raises $30 million in funding from Prosus for its AI call screening assistant that supports 10 languages. The app currently handles calls from unknown callers, but it has ambitions to handle calls from known contacts to know the reason. The startup also wants to make outbound calls with its app to book appointments. All of this is very ambitious in India’s multilingual landscape, where not all things are digitized.

• Rylo (formerly Nagish), the platform that helps people with hearing impairments, raised $85 million in funding led by Canaan. The round also has contributions from General Catalyst’s Customer Value Fund (CVF) in the form of growth equity. With the new funding, the company aims to add workplace accessibility, network improvements, and real-time fraud detection to its app.

• Krisp dropped its Voice Translation v3 model with the ability to audit calls live for admins. Plus, the new model has industry-specific vocabulary built in, and it is easier for admins to add custom phrases.

• Deezer has been battling against AI music hard. The company has been offering a tool to other streaming platforms to identify AI-generated music. Last week, the company released a tool for users to identify such music using their playlists imported from other platforms like Spotify and Apple Music. The tool supports detection in 27 languages across 20 platforms.

• Warner Music Group acquired Sureel AI, a tech company that can track how AI is using existing tracks by artists. For labels, AI has presented a new question about how to determine whether the tracks it owns are being used. AI’s usage of data fed for training is fuzzy. There isn’t a standard way to say that a specific percentage of a track was used to generate a new one. But any way for labels to measure usage could be critical for business.

Partner Spotlight with Atomik Growth

In a chat with a16z, ElevenLabs CEO Mati Staniszewski talks about how he and his co-founder, Piotr Dabkowski, got inspired to solve for badly dubbed movies and created a company that thinks that voice is the new interface for human-computer interaction.

Sponsored content

Thank you for tuning in. Keep listening.

This newsletter is by Ivan Mehta, a freelance reporter at TechCrunch. It covers AI and technology in voice, audio, and music. Email: voiceaiweek@gmail.com or im@ivanmehta.com

I am starting this newsletter with this section because I have a lot more to say about a particular topic than the news in the past week: translation. I spent the last week in China going through multiple cities, including Shenzhen, Hangzhou, Suzhou, and Shanghai.

In the past few years, all of us have read about the transformative quality of large language models that can understand multiple languages. All AI labs flaunt their benchmarks while releasing a new model. But in practice, I didn’t see the effects of the new AI model when it came to the translation of text or speech.

When I was in China, I had to use apps like AliPay and WeChat a lot. Often, tea and coffee shops in China would have a mini app inside these apps that can help you order. There is a floating translation option in Alipay, but the success rate of that working and working well was not great.

I was in a lot of briefings and Q&A sessions where an executive was speaking Chinese. When I used iFlytek’s S6 recorder, the audio captured was good, but the on-device translation was very average, and its interface was janky to look at and follow the speech at the current point.

AirPods’ translation feature only worked well when I was in a one-on-one conversation, and that too missed the mark of what I had asked and what the other person was saying.

In China, Google’s services are restricted, so I relied on Apple’s translation app to take photos and convert text on menus and signs to understand the basics. The translation through the camera is not great, but it is good enough to make do and place an order at a restaurant. With WWDC scheduled today, I wish that, beyond Siri, Apple would add more intelligence to translation. The only bit that worked well was holding the phone with the translation app to the other person and conversing with them.

I was trying out Even Realities G2 Glasses, and the translation was very good. It worked well when an executive spoke in Chinese, and I was looking at the text on my glasses, but a lot of times, that felt like I was catching up to the speed of text that was whizzing past my tiny screen.

But there are a few practical caveats to this. You need to have the glasses on you and connected to your phone. And you need to activate the translate mode when conversing with someone. To show them your side of the conversation, you need to open the app. Plus, with this particular pair of glasses, you can’t translate real-life text as there is no camera.

There are a few challenger brands like Inmo that offer both camera and screen with a translation feature, but I just experienced them in a demo at Global Connect Show (GCS) and don’t have a verdict to pass yet.

With AI advancing so much, I think there should be better standalone apps and in-app translation features for travelers. We can’t stick with old clunky interfaces that barely work and get translation good enough for users to understand a word or two and make a purchase decision at a shop or restaurant. With AI, I feel that the act of translation should feel more fluid, reliable, and easy to use. Maybe work on a better translation before chasing AGI?

Top News

Suno scores $400 million in new funding at a $5.4 billion valuation

Music creation app Suno said that it has raised $400 million in a Series D fundraising led by Bond Capital with investment from IVP, Forerunner, Union Square Ventures, Alkeon, and Quiet.

The company’s latest funding comes after its $250 million round in November 2025. Since then, it has settled its lawsuit with Warner Music Group (WMG) and seen a bump in valuation from $2.45 billion to $4.5 billion.

The legal trouble still exists in the form of lawsuits with Sony Music and Universal Music. Recently, both labels added more than 61,000 songs to the lawsuit, claiming that they were used in model training by Suno. This increases potential damages for Suno if the court favors the labels. However, Suno is urging the court to let it keep the number of songs used in model training under wraps.

However, more people are using the app. The company said in February that it has more than 2 million paid subscribers and $300 million in annual recurring revenue. The company also reached #1 in the music category in the App Store. The company said in its announcement

In recent months, we’ve seen Suno become part of culture in ways that continue to surprise us. Family members are turning text threads, group chats, and inside jokes into songs. People are writing songs for birthdays, graduations, and even work events. Viral trends helped propel Suno to #1 in the App Store’s Music category in dozens of countries.

It is not yet clear if people are using this for fun or if there is a trend of professional users using the app for production or ideation. Investors are betting on the usage going up and also the company settling its lawsuits.

In the past few months, ElevenLabs, Google, and Scale AI have released license-safe music models. Suno is seemingly going in the same direction, as it said it plans to release its first music model “developed in partnership with the music industry.”

Quick Bytes

• Sesame, the voice AI startup from Oculus founders, launched its app in public preview. The app now has four agents, Maya, Miles, Simone, and Charlie, with their own memories and personality types.

• AethexAI, a startup concentrating on Africa and the Middle East, raised $3 million in funding and opened up its platform. The company also released its Kora series of models that work with localized dialects of English, French, and Arabic.

• Google is going to roll out on-device scam detection for instances where the miscreant is using number spoofing and calling you from what seems like a number on your contact list. As CNET noted, the caveat is that this only works when both parties have the Google dialer, contacts, and Messages app installed.

• Customer support startup Fin (formerly Intercom) released a new model called Apex Flash. The company said that the new model is almost 20% faster than the previous version for time to first audio. Plus, it also improved the resolution rate by 24%.

• Google released Gemma 4, a 12B model that doesn’t have a separate encoder for audio and video inputs. With this, the company released its Edge Gallery app and Edge Eloquent app for dictation on Mac, which use local models for their features.

• Two years after pulling back its voice-led ordering system, McDonald’s is piloting a new one in partnership with Google. An account called McFranchisee on X noted that in its test phase in five stores, the system has processed over 1 million transactions.

Off Topic

One thing I always miss after my trips abroad is beverages and snacks that I get in convenience stores. During my trip to China, I became a fan of both iced black tea and Chrysanthemum tea, which you get in 7-Eleven or Family Mart stores. I also liked the dark chocolate KitKat that is not available in India.

I did see cucumber-flavored Lay’s chips that I couldn’t try. A chewy version of Alpenlibe was also a new discovery for me. I also brought a ton of new flavors of Pocky back home.

Partner Spotlight with Atomik Growth

In a chat with a16z, ElevenLabs CEO Mati Staniszewski talks about how he and his co-founder, Piotr Dabkowski, got inspired to solve for badly dubbed movies and created a company that thinks that voice is the new interface for human-computer interaction.

Sponsored content

Thank you for tuning in. Keep listening.

This newsletter is by Ivan Mehta, a freelance reporter at TechCrunch. It covers AI and technology in voice, audio, and music.

Email: voiceaiweek@gmail.com or im@ivanmehta.com

May's funding was all about big bucks

May was a strange month for voice AI-related funding. The number of deals was visibly fewer, but some mega deals propelled it to the top of the funding chart for this year, with over $1.8 billion poured into startups related to voice.

Two deals stand out here: Sierra’s $925 million and Hark’s $700 million “seed” round for its secretive product.

Sierra’s funding is in line with previous months’ signals around rising demand for customer support and experience startups. The company reached a valuation of $15 billion with this round and has been serving 40% of the Fortune 50 as customers.

Hark’s funding is curious because investors are betting on future promise. We haven’t yet found a perfect formula for a personal assistant that works well. But the idea is worth chasing after.

Vapi’s $50 million deal was a pure voice AI play this month, with the company reaching a $500 million valuation. The startup said it handles 1-5 million calls a day now.

Sector-wise, healthcare was a strong sector with deals of Commure, Basata, Enzo Health, and Kin Health totaling over $120 million in funding.

Signals & Experiments

India is a big voice market, and that was evident from demos at the Mumbai Tech Week, where the likes of Urban Company and Meesho showed their voice solutions for customers for use cases like support and item discovery. While I wasn’t present on day two, Indian AI startup Sarvam was slotted to demo its own tech. Even the Maharashtra Government talked about voice conversations in Marathi for its farmer-facing AI app.

Some people I spoke with noted that when they received a call from AI, their experience wasn’t great. The fact that they knew it was AI on the other end didn’t work on their part. And in some cases, the conversation didn’t flow naturally.

In an on-stage chat with me, NPCI (National Payments Corporation of India) head Dilip Asbe, who looks after the Unified Payment Interface (UPI) standard, mentioned that while tech could be a viable interface for engaging users, the payments body is still identifying the right use cases, and it is still early days to deploy voice. This means there is still work to do to engage users effectively with voice AI both on the enterprise and AI front.

Model Behaviour

ElevenLabs launched two new models this week: a new music generation model called Music v2 and dubbing called Dubbing v2. Both had a common focus on bringing more depth to the output and also giving more control to creatives.

The music model’s biggest feature is its ability to change genres mid-song. For professional users, the company has added a better understanding of song structures and building blocks like verse, chorus, and bridge.

This allows for track creation via structural prompts rather than manually stitching short clips together. The latest model also lets users select a particular part of a song and prompt it to change it without affecting the rest of the song.

ElevenLabs said that the Dubbing v2 model supports over 90 languages while preserving the tonality and emotions of the original speaker.

The company is making business moves with this release, too. First, it is inviting creators to try out the dubbing product at a discounted rate. Second, for studios and broadcasters, it is partnering with human translators, expert voice casting, and professional audio mixing experts for localization services. More like, forward-deployed creators.

While ElevenLabs’ big chunk of revenue still comes from its enterprise business, the company is making sure that it plays a big part in the creative process. The company is competing with the likes of Google, Suno, and Stability AI in music, but it is also covering other workflows, such as dubbing.

Quick Bytes

• Meta is building a voice recording pendant, as per a report from The Information. But that is hardly surprising given the company acquired Limitless, which also made pendants. The question is: Would it make people trust a Meta-made pendant?

• One of the key factors for a model to understand someone’s voice is to isolate it from noise. Hardware company BlueParrot has debuted two new headsets for long-haul truck drivers that use voice to remove noises that you hear when someone is driving.

  

• In Korea, a major case has unfolded where a YouTuber is accused of using AI to manipulate text messages and voice recordings of a movie star to defame him and accuse him of dating an actress when she was a minor. We will possibly see almost TV show and movie-like cases in the future where voice and video manipulation might be key factors.

• London-based company Voxmind raised roughly $734,000 for its voice deepfake detection tech. The money is small, but the technology is critical, especially for enterprises. The startup is entering a space where the likes of AWS are retiring their ID services, creating a good opportunity.

• Sony expanded its lawsuit against music AI companies Suno and Udio. The label added 61,026 songs to the Suno case, expanding the ceiling of damages to over $9 billion. It also added over 30,000 songs to the Udio case.

Off Topic

I’m in China this week looking at new gadgets, and I am particularly excited to see how much AI is used in these gadgets. I haven’t seen any demos yet, but I am fascinated by this robot in the hotel that I am staying in, which could deliver small items and some food to your room without human intervention. This is cool because it eliminates the problem of “we don’t have staff available” right now. And also possibly allows workers to rest a bit more.

This is also a good ground to test out translation tech, especially when it relates to real-time voice and text translation.

Partner Spotlight with Atomik Growth

In a chat with a16z, ElevenLabs CEO Mati Staniszewski talks about how he and his co-founder, Piotr Dabkowski, got inspired to solve for badly dubbed movies and created a company that thinks that voice is the new interface for human-computer interaction.

Sponsored content

Thank you for tuning in. Keep listening.

This newsletter is by Ivan Mehta, a freelance reporter at TechCrunch. It covers AI and technology in voice, audio, and music.

Email: voiceaiweek@gmail.com or im@ivanmehta.com

Google’s voice interaction push and Spotify’s voice generation push

I. Google IO was packed with announcements about AI agents. However, a sub-theme was voice interaction across different apps and modalities. These features might excite people who already use voice for dictation and prompts. In isolation, they look good, but for a company whose services are so widespread, the new voice announcement seemed fragmented. I’ll break down what I mean by that as I go along.

One core idea Google showed off was that now you can talk to its Workspace apps like Gmail, Keep, and Docs. But the function of voice in each app was different. In Gmail, it was more about providing a fuzzy context to the assistant and letting you find an email; in Keep, it was to polish up transcription and post a formatted chunk of text as a note; in Docs, it was about interacting with an assistant that can take multistep commands, including fetching information from sources like email and files in Drive. But are these all one assistant? Do they carry the context from one place to another? It is not clear.

WSJ reporter Nicole Nguyen tested an early version of the Docs Live product, and let’s say there are still some things that need to be ironed out before the feature goes into production.

Google’s product strategy has always been confusing, and the company often duplicates features and products. Earlier this month, it released an AI-powered dictation feature called Rambler, which cleans up speech. Now, a similar feature exists in Keep. Maybe the idea is that if I don’t use Gboard, I still get to use the built-in dictation power. But what takes precedence when there are two features?

There are some places where Google didn’t talk about the voice feature out loud. For instance, its agentic coding platform Antigravity now has voice support. Plus, the company also showed off voice commands on its Gemini Mac app that followed instructions and got tasks done. We don’t know if the latter is going to make it to a wider release.

One surprise was Google announcing its “Audio glasses” that will arrive this year. These glasses isn’t designed just for music listening, but they are akin to Meta Ray-Bans and don’t have a screen as a display. The company is developing these glasses in partnership with Warby Parker and Gentle Monster.

With these glasses, users can fire up Gemini to get information or complete a task like ordering a coffee. The glasses can use cameras for visual context to translate text or speech, give you directions based on what you are seeing, and read out details about the object in front of you. Plus, you can snap pictures and ask Gemini to edit them for you. A lot of people use Meta’s glasses to take photos, but a few of the AI features Google listed could be useful if they work in real-life conditions.

II. Spotify also had a big voice AI-focused announcement week with a bunch of new products and features unveiled at its investor day event. Google’s whole spiel was about talking to AI, and Spotify’s concentration was on generating voice content through AI.

The company’s marquee announcement in this area was personal podcasts, a way for users to use a prompt for a topic to generate a new podcast using AI. You can prompt to create a daily or a weekly briefing about a current affairs topic or a sports team that you are interested in, or generate a one-time podcast to understand a concept.

Other products like Google NotebookLM, ElevenLabs Reader, and Adobe Acrobat allow for a similar podcast creation for personal usage. There are apps in this area, too. Former NotebookLM developers tried building an app called Huxe, which is shutting down. Another app called Sun is working in the same modality under a16z Speedrun. Spotify understands that more people are trying to generate AI audio personal entertainment and information use cases, and wants to be at the center of this.

The company also released an experimental desktop app called Studio by Spotify Labs, which will connect to your email, calendar, and notes to create audio briefings about your day or a trip. Ideally, you don’t need a separate app for that, but Spotify seems to have a larger ambition to be in the productivity space. The company hints at this in the app description:

With your permission, it can take action on your behalf: researching topics, using a web browser, organizing information, and helping complete tasks. It can also connect to the tools you use every day, like your calendar, inbox, and notes.

Every company wants to build agents that take action on users’ behalf. Spotify might not be there yet, but it can test the waters with its new app. Given its desktop presence, Spotify could easily venture into other modalities like meeting recordings.

Spotify is not stopping at personal audio generation, though. The company released two other products for public consumption. The company is partnering with ElevenLabs to let authors create audiobooks using AI voices. It also struck a deal with Universal Music Group (UMG) to let users produce AI remixes.

Then there are discovery features to get to the podcasts and audiobooks using AI. If you have questions about an episode or a concept, you can also ask the assistant about them rather than going to Google or ChatGPT. Spotify is loading up the app and tools with AI to get to every kind of audio possible, but this might end up confusing users.

Signals and Experiments

April was an interesting month for voice AI funding. The dealflow amount was higher than that of March, but there was only one deal over $100 million, signaling mid-sized cheques.

Deals in the customer service vertical keep on flowing, but we saw that HR/recruiting, along with healthcare, are strong sectors for investors to fund startups.

April was also the first month in the year when no new unicorns were minted.

While April was good, May seems to be shaping up as the blockbuster month with two $700 million plus deals.

Quick Bytes

• As the market for AI-generated personal podcasts heats up, former NotebookLM developers’ app Huxe is shutting down. The company didn’t provide core reasons, but said that the team will move on to other projects.

• The creator of the newly open-sourced transcription model, Mega ASR, claims that it performs better than Gemini 3 Pro and Qwen 3 ASR.

• LG has partnered with Kardome to add voice tech in its TVs that allows users to control the device with voice, even when audio is playing loudly. We definitely need more tech that works in loud environments.

• ElevenLabs said that voice creators have earned $22 million through its products, up from $11 million cited last year. The company is adding more than 200,000 premium titles to its ElevenReaders Library. It said that subscribers who pay $11 per month can get 20 more hours of listening.

• Stability AI released a new family of music creation models, called Stable Audio 3.0. The small models are designed to make sound effects and music of up to two minutes locally on a device. The larger models are designed for professional use cases with the ability to conjure up a six-minute-long track.

• Suno is in more legal trouble as the indie band The American Dollar filed a lawsuit against the startup, alleging that it used 236 of their songs to train AI models. The band said it has suffered a nearly 80% loss in licensing revenue due to AI copying its style.

• Brett Adcock’s new venture, Hark, a personal AI platform including software and hardware, got a massive $700 million in funding. The company said it is developing an assistant that uses speech, vision, and memory to interface with users.

• ElevenLabs said that it will now use Google’s SynthID to watermark AI-generated audio. Meanwhile, TikTok also said in context of its UMG deal that it plans to ramp up its efforts to remove unauthorized AI music from its platforms.

Partner Spotlight with Atomik Growth

In a chat with a16z, ElevenLabs CEO Mati Staniszewski talks about how he and his co-founder, Piotr Dabkowski, got inspired to solve for badly dubbed movies and created a company that thinks that voice is the new interface for human-computer interaction.

Sponsored content

Thank you for tuning in. Keep listening.

This newsletter is by Ivan Mehta, a freelance tech reporter at TechCrunch. This newsletter is an attempt at covering what is happening in the industry of voice, audio, and music.

Email: voiceaiweek@gmail.com or im@ivanmehta.com

Google makes an AI dictation move as Wispr’s valuation balloons

Google announced a new AI-powered dictation feature called Rambler at its Android Show: IO Edition event last week. The feature lives in Gboard, and users can access it in any app to dictate speech and get clean text free of filter words.

Google already released an experimental app called AI Edge Eloquent last month on iOS with a focus on offline models. However, there was no keyboard, and you had to manually paste the dictated text into another app. Rambler is a more polished and productized version that is integrated with Gboard and will be accessible to Pixel and Samsung Galaxy users to start with.

The first thought in the minds of many users would be that this is bad for apps like Wispr Flow and Typeless. But I have a few counterpoints. Dictation apps are still a desktop-heavy category. These apps have their mobile counterparts, but people use them because they use the desktop version, and possibly want to carry over the same experience, along with their custom dictionary.

Rambler might help users get acquainted with AI-powered dictation. But if anyone seeks advanced features, they will move to other apps.

This is where the discussion about AI-powered dictation being a feature, product, or company starts. Google saw that many people are using apps in this category and decided to build a feature, but that doesn’t mean power users will automatically gravitate towards it.

The trend is evident in another news story this week: Wispr AI is seeking to raise $260 million in funding at a $2 billion valuation, up from $700 million in the last round, as reported by Bloomberg. Menlo Ventures, who are admirers of the product and led its Series A, will reportedly lead the round again, as per the report.

The app is popular with venture capitalists and folks working at companies like Nvidia, Amazon, Groupon, and Mercury. And as we have seen in the case of Granola, if certain people in tech feel that they can’t do without a particular product, the valuation rises.

Fish Audio vs voice artists

For the last few days, many voice-over artists on LinkedIn have been complaining about an AI platform called Fish Audio, saying that their voices ended up on the platform as models without permission, with many of them being used thousands of times.

Fish Audio lets users upload voices and have others use them for voice cloning and AI-powered reading/narration projects. Artists mentioned that this model has allowed others to upload their voices without their consent. The platform has a takedown process for requesting the removal of audio. But several artists complained that the process is tedious, and it took them days to complete the takedown request.

“I’m sorry this happened. The reported community-uploaded model has been removed. Fish Audio is a UGC platform, and we take reports of impersonation or rights violations seriously. In the past month alone, we’ve removed over 100 reported or policy-violating community voice models as part of our ongoing enforcement efforts,” Fish Audio co-founder and CEO Rissa Cao said in a reply to a voice-over artist’s post.

Copyright is a massive issue in the voice world. Several artists like Taylor Swift and Indian singer Arijit Singh have moved to protect their voices. Lawmakers are also looking at this problem. Denmark passed a law this year that protected the voices of its citizens from AI cloning. In 2024, the state of Tennessee passed the ELVIS (Ensuring Likeness Voice and Image Security) Act with similar protections.

Signals & Experiments

The mobile dictation dilemma

Apps are realizing that voice is a growing input method for a lot of users. Voice input is particularly useful when users need to dump in a big chunk of text. Email is one target use case, and increasingly, the second one is AI apps for prompts.

I have had mixed experiences with these tools. I very regularly use this tool called Littlebird, an assistant that knows my work context by “reading” my screen. I have tried using its own transcription tool for prompting, but it often falls flat. Plus, on desktop, I have the option of using apps like Wispr Flow or Willow in any app easily.

It’s a different ballgame when it comes to mobile experience. As I have written before, Apple is breaking the AI dictation experience with a new flow in iOS that requires you to swipe back to the app you were typing in after activating a voice session. So when email apps like Avec use their own dictation engine, it might make life easier for people. Plus, as dictation is one of its core features, the company will keep working on improving quality and personalization.

Model Behaviour

Solving for Duplex and Expression

Labs are edging towards making voice interactions more human. AI labs are focusing on two main fronts: turn detection and interruptions, and voices sounding more expressive.

Mira Murati’s Thinking Machines Lab demoed a full duplex multimodal model that can listen and react at the same time. The demo showed real-time interaction, real-time translation, and a model triggering a response when a person walked into the frame.

Resemble AI released a new model called Dramabox, which lets you describe a character and then have the model enact the voice using a script. The model is doing this without any specific expression tags, which is kind of impressive.

Quick Bytes

• OpenClaw creator Peter Steinberger said that he has built agents that listen to his (and his team’s) meetings and proactively start working on tasks like creating pull requests based on the discussions. This is a goal for a lot of meeting note-taking companies, but they might not have the level of access to the tool they would ideally hope for.

• A study in the UK suggested that 40% of general physicians used ambient AI scribes to transcribe patient sessions and take notes. This study comes after the National Health Service (NHS) laid out guidelines around using the technology.

• Musician and producer Jack Antonoff, who has worked with artists like Taylor Swift, Lorde, Lana Del Rey, and Kendrick Lamar, lambasted the use of AI in music creation.
  
  “So to everyone who is gassed up about the new ways you can fake making art, by all means, drive right off that cliff. We’re genuinely happy to see you go,” he said in a post on his Instagram account.

• Meta is finally on the conversational assistant train. The company demoed a voice chat with Meta AI, powered by its Muse Spark model. Users can point the camera at different objects and ask questions about what is in the frame using voice.

• Indian investors had a good week. PeakXV led voice orchestrator Vapi’s $50 million round, which took the startup to a $500 million valuation. Meanwhile, growth stage firm Activate invested in ElevenLabs through a special vehicle. The firm will help the voice AI giant with India expansion.

Off Topic

I started testing Project Mirage’s Dune gadget, which has three buttons and can change its function based on the app that is open. I haven’t quite gotten to making custom controls for it, but right now, the biggest use case for me is to paste URLs into Slack, and also use mic and video controls.

Partner Spotlight with Atomik Growth

In a chat with a16z, ElevenLabs CEO Mati Staniszewski talks about how he and his co-founder, Piotr Dabkowski, got inspired to solve for badly dubbed movies and created a company that thinks that voice is the new interface for human-computer interaction.

Sponsored content

Thank you for tuning in. Keep listening.

This newsletter is by Ivan Mehta, a freelance tech reporter at TechCrunch. This newsletter is an attempt at covering what is happening in the industry of voice, audio, and music.

Email: voiceaiweek@gmail.com or im@ivanmehta.com

Plus, there is a question of retaining data and using it to plug into AI tools as well. Some tools might not allow that if you’re not on the business plan. One of the answers is Talat, an offline and private notetaking app that costs only $49 one-time purchase.

The app works on your system, both Mac and Windows, and uses system audio for transcription — just like Granola. It doesn’t require you to create an account, promises regular updates, and also allows you to record audio if needed. It shows you meeting transcripts and an AI summary once the meeting is over, and you can use text search to query meetings.

All models for transcribing and summarizing meetings are local LLMs, and you can choose to customize them. You can also change the prompt for summaries if you need to switch to another template.

Last month, Granola locked local transcripts and the breaking AI agents. There was a lot of debate at the time around data ownership. Talat wants users to own their own data, and offers exports and MCP connectors to use in other AI tools.

Talat’s backstory

Talat is built by friends Nick Payne and Mike Franklin. Payne, who has spent a large part of his 20-year career in web development, said that when there were a lot of people on the call, and he had to lead the meeting, he felt anxious. A few years ago, someone introduced him to Granola, and he thought the app was great and allowed him to focus on meetings while taking notes.

Sometime ago, Granola started restricting people's free access to only 30 days of meeting transcript access. And this switch was sudden.

“I started getting pop-ups in the app saying, ’ You’re going to be downgraded, and you’ll lose anything more than 30 days of your meeting history. And as someone who’s run lots of SaaS tools in the past, I knew that they have to make money, but it’s so arbitrary that they are not letting me get my 30 days of meetings, especially because they’re cached on my machine. I can literally see them,” Payne said.

He also wanted control of his data and wanted to store everything locally, so he started building Talat.

Technical stack

Payne was fascinated by Granola’s design and function and started digging into different voice models. He built a companion app, which helped prep for a meeting using voice. The tool also gave him tips during an ongoing conversation. But he felt that he couldn’t solve the latency problem to get instant answers.

While researching voice APIs and libraries, he came across Apple’s Core Audio Taps API, which could be used for system audio recording. Based on that, he built a more friendly abstraction layer called AudioTee, which makes it easier to use poorly documented Audio Taps.

He also used the Fluid Audio framework that lets you execute AI functions using CoreML. This enabled Payne to also do streaming transcription. This also allows for functions like speaker identification. Plus, the whole stack enables offline transcription through models like Nvidia’s Parakeet.

“I’m realizing, as we’re building Talat, that privacy and offline capability are huge features. Being able to have your meeting notetaker provide real-time transcription without an internet connection is a great unlock,” he said.

Payne understands that Talat won’t be able to cater to all kinds of workflows. With the app offering exports and MCPs, he said that he wants to design Talat in a way to “get out of your way” and let you build your own workflows.

In the past few months, a lot of open-source or one-time cost tools have cropped up, including OpenOats, Museli, and SamScribe. The idea is to let people who might not need a lot of features besides having transcription handy have control over their data.

All voice roads lead to customer service

A lot happened in the past few days:

• Bret Taylor’s Sierra raised a massive round of $925 million.

• Customer service company Netomi raised $110 million.

• OpenAI released new models for real-time conversation and translation.

• ElevenLabs went past $500 million in ARR and announced that venture arms of enterprises like Salesforce, NVIDIA, and Deutsche Telekom had become its investors

• Airbnb said 40% of its customer support queries were handled by AI

• Bajaj Finance said that 30% of its outbound calls in the last financial year were handled by voice agents

We have seen multimillion-dollar fundings, new enterprise partnerships, and companies claiming that AI is handling more of their customer queries. All of these points point in one direction: customer service is voice’s biggest sector. Calling is an old channel, and companies are racing to digitize it. Sierra’s announcement captures it well.

“We’re in the very early stages of AI adoption. Agents have enabled us to digitize the last remaining analog channel: the telephone. All of us can now get things done without waiting on hold, in any language, and at any time of day (good riddance, IVR systems).”

Even in ElevenLabs’ new post about welcoming enterprises and celebrities as investors, the focus was on building and deploying voice agents.

Voice companies and investors are in sync that while there are parts of the technology, like transcription and dictation, useful to consumers, the real money is in reducing call center costs and increasing sales inflow for enterprises. At the same time, for small businesses with less than a 5-person support team, customer service companies can create AI tools that can do some preliminary conversations without human intervention.

While AI voice agents aren't necessarily cheaper than human labor yet, they offer unparalleled scalability, 24/7 availability, and the ability to maintain consistent context across thousands of simultaneous calls.

For a few years, we are bound to see companies boasting about how their systems are handling more customer queries without human intervention. However, there are multiple surveys indicating people want to talk to human agents.

The tech industry is not thinking about adding more humans despite what people say in these surveys. It is striving to make voice AI agents “more human,” by improving voice generation, cleaning audio, or passing more context.

There is a human cost to this. Support jobs are being impacted as AI is taking up some of the work. Gartner predicted that this staff cut will be messy as companies slashing support jobs will need to rehire people in 2027 under different job titles. This adds a lot of uncertainty among workers as they might not want to stick to the same industry after being laid off.

At the moment, there is no stopping this AI CX momentum as investors are pumping money into this sector. As per analytics company Tracxn, customer support startups have raised more than $2.47 billion in funding this year.

Signals & Experiments

I have been testing the new reMarkable Paper Pure tablet for the last 10 days. The tablet has a monochrome screen, a stylus that feels like writing with a pen on paper, and an on-screen keyboard. This is a great device for notetaking or even getting long drafts of blogs started with no distraction.

While the writing is great, I wondered if having voice-based notetaking would make the device better. Putting a mic into the stylus might drain the battery and make it a more complex design. It might be weird to put a mic in the tablet itself because who holds a tablet near their face to take notes?

The best bet would be software integrations that would allow me to integrate my voice or notes apps to reMarkable’s system to import notes, and that might be happening soon.

During my briefing, Paper Pure, the company, had said it would look into importing meeting transcripts or notes to the tablet.

“We know that now you know Gemini, Copilot, and there are many things, sort of generating voice transcripts and then being able to put it in the context of whatever you’re sketching or whatever you’re doing is super important.

Quick Bytes

• Netflix is testing a new voice AI feature, as reported by Janko Roettgers of Lowpass Newsletter. The company is letting users press the Netflix button on their remote and search for movies and shows with phrases like “I need a good cry,” “watch in the background,” and “help me stay awake

• Spotify added a new GitHub-based tool that lets users create podcasts from coding tools like Codex and Claude Code and import them to their library. There are tons of ways to now create AI-powered podcasts. With this move, Spotify wants to make sure it is the choice platform for users to listen to any kind of audio.

• Agent building platform Voagent was acquired by customer service platform Aircall. We might see more consolidation moves in the customer service and support industry in the coming months.

• Music AI had an interesting week. On the one hand, Suno is aiming for a $5 billion valuation for its latest funding round, as reported by Axios. On the other hand, rival Udio admitted that it scraped YouTube to train AI models in a case with Sony Music. While these music services are growing, copyright cases could be a big roadblock.

• Data provider Appen said that it is teaming up with Hugging Face to provide datasets for the Open ASR leaderboard. These datasets will include data to gauge word error rate (WER) across different situations, including evaluating speech recognition for average U.S. vs. non-U.S. accents. Audio benchmarks are fragmented, and it is hard to have a view on which model is superior, with a few benchmarks that can cover only a few situations. This is a welcome step for independent and varied voice testing

• Voice-based scams are getting more commonplace as miscreants are using deepfaked voices of relatives of their target or public figures. Though security experts often advise that “don’t pick up unknown numbers,” it doesn’t work in real life, as you might be expecting a call from a person or service whose number you don’t have in your address book

• Besides adding new institutional investors, ElevenLabs also closed a $100 million tender offer, months after it floated one for employees in September 2025. The company also added artists like Eva Longoria and Jamie Foxx to its investor list. For the voice AI company, this could be a play to secure IP and make Hollywood connections to integrate deeper into the creative community

Subscribe now

Off Topic

I always look for apps to store links or articles, especially since Mozilla decided to shut down Pocket (RIP).

One app that I have tried out and written about in the past is former Twitter engineer Joe Fabisevich’s app Plinky. It is great, but limited to the Apple ecosystem, to save links beyond articles and categorize them. Fabisevich recently launched a Mac app in beta to make it easier to save links from the desktop.

Another option for cross-platform saving is Folio, which is made by the former head of product for Pocket, Nick Chapman. The app is available on iOS and Android as a native app, and there is also a web app and a Chrome extension for desktop use. Overall, a very clean app.

Partner Spotlight with Atomik Growth

In a chat with a16z, ElevenLabs CEO Mati Staniszewski talks about how he and his co-founder, Piotr Dabkowski, got inspired to solve for badly dubbed movies and created a company that thinks that voice is the new interface for human-computer interaction.

Sponsored content

Thank you for tuning in. Keep listening.

Email: voiceaiweek@gmail.com or im@ivanmehta.com

This newsletter is by Ivan Mehta, a freelance tech reporter at TechCrunch. This newsletter is an attempt at covering what is happening in the industry of voice, audio, and music.

While Deepgram offers full-stack services, it is revered for its speech recognition models. The company says companies have built over 2,000 products on top of its voice stack. I sat down with the company’s CEO, Scott Stephenson, to talk about its new Flux multilingual model release, context carrying for voice agents, and why it is difficult to build open-source, speech-to-text models.

Flux multilingual

Last year, the company made a shift in its model with a new Flux series, focusing on powering real-time conversation instead of batch processing.

The earlier Nova series of models is still active, but it wasn’t designed from the ground up with live conversation in mind. As more enterprises shift to AI-aided consumer service and experience, demand for voice agents is rising, and voice AI providers are racing to build models and infrastructure that solve problems in real-time conversations, such as interruptions and audio cleaning.

In multiple countries, people speak in a mixed code manner. That means they switch between languages mid-sentence. In April, Deepgram made its Flux series of models multilingual, with a focus on ten languages. The company has built conversational nuances such as turn detection and code switching in the Flux model.

Stephenson thinks that more companies are going to switch to Flux given the high demand in the consumer support world.

“Nova was a prerecorded and batch-first architecture that we also made real-time. It was more for real-time, closed captions, and that type of thing. So less for hyper-performant voice agents with the lowest latency and that type of thing,” Stephenson said.

“Flux is probably more performant, is probably faster, and it certainly is more performant in the task of if you’re switching languages and noticing that’s happening and then having a higher quality, higher accuracy when that’s happening.”

He said that typically models looked for silence to detect turns, which wasn’t the best method because the user might have gotten distracted or was trying to remember something. If I were recalling my account number, I don’t need a voice agent to interrupt me.

Deepgram said that the company needed to include different filter words and cadences of different languages to make the model truly multilingual.

Passing context in voice models

Deepgram CEO noted that the way you make voice agents’ voice context will matter as much as having high-performing models. For instance, if your voice agent can detect that someone is in a loud environment, and it speaks more slowly and louder, the person on the other end of the call would be able to understand the voice AI bot better.

Stephenson said that in the current voice agent paradigm, where there is an architecture of speech-to-text, text-to-text, and text-to-speech models, the more context-like turn detection and pauses that you pass to models, the more accurate and “human-like” the agents become. He said that cues like a change in tone are critical to detecting if someone is ending their turn.

“This is one of the biggest changes that is going to happen in this next evolution and revolution in voice agents, where you add context everywhere. And this is what the flux line is about. Flux conversational speech recognition adds more context to perception. And later on, with more flux releases, we’ll talk about adding context to the rest of the stack,” he said.

“I think about this is that there are sort of two things that you’re fighting when you’re building a voice agent. Between turns, you don’t want to wait 10 seconds of silence until you finally kick off and start responding. Then you have the other side, which is if you look for just a little bit of silence, and any little bit of silence you find, you jump right in. Then everybody is like, ‘This thing is interrupting me all the time,” he said.

Future of the open-source model

Stephenson says when the cost to get started in voice drops, we might see less activity on the model front in open-source. But we might see more activity in people building code for adding context or orchestrating voice agents.

He thinks that it is hard to build a good open speech-to-text model.

“It’s hard and expensive to do open-source speech-to-text models. There are way fewer people doing that. Whisper was launched, I don’t know, three years ago, three and a half years ago, and that probably still is the best open source speech-to-text engine. So there isn’t a lot of activity in that area,” he said.

He thinks the reason for that is that someone building a speech-to-text model has to get millions of hours of speech data to work with and have money to train models.

Notably, labs like Mistral and Cohere have released open-source transcription models recently, but they do have financial backing to spend on research.

He thinks that it is easier for builders to get into open-source voice generation and orchestration than transcription.

Way forward for DeepGram

The company has different lines of models for users, but it wants to give them the option for “good, better, best” to choose from in different parts of the voice stack. Just like ChatGPT offers Instant, or Google offers Flash, Deepgram wants to offer a cheaper and faster version of its top models.

In an interview with Week in Voice AI, Deepgram investor Zach DeWitt called “Google of voice.” I asked Stephenson about the comparison; he leaned into the cultural parallel, noting that Deepgram’s work environment allows people to expand their knowledge.

“Google has an amazing work environment. So when you’re at Deepgram, there are sort of infinite possibilities to expand your knowledge and move from one area of the company to another and engage with different folks and have ample time that’s outside of your scheduled work that you’re supposed to be doing in order to expand and explore,” Stephenson said.

ElevenLabs acquires a team from a Polish voice AI startup to bolster research

ElevenLabs continues to grow in revenue and valuation, but it also wants to remain, at its core, a research lab around sound and audio. The company’s value came from its ability to create unique, expressive voices. Now, it wants to accelerate research in voice generation by acquiring a team behind a Polish startup called Papla.

Papla was founded by Hubert Siuzdak, Jakub Abramczyk, Dominik Stachura, and Tomasz Dentko.

The startup largely focuses on voice generation with an API that provides realistic voices in English in different accents. The company claims on its website that you can clone a voice in just 10 seconds of audio. The company’s model ranked higher than models from ElevenLabs and Cartesia on the TTS Arena V2 leaderboard.

At the core, research, and the Polish connection, which possibly attracted ElevenLabs to hire the team.

The key achievement of the startup comes from Siuzdak’s research. He wrote a paper on solving for natural-sounding voice generation with better pronunciation through a two-staged architecture. His second research was around making a fourier-based vocoder output high quality while using fewer resources. His third research was focused on using low bitrates such as 0.98 kbps for speech and 2.6 kbps for music for high-fidelity output.

All this means that ElevenLabs can chase top-end research problems that could eventually result in products.

Deepgram’s new speech recognition model focuses on multilingual code-switching use cases

Fresh off the $130 million funding in January, Deepgram released a new speech recognition model, or rather, a multilingual version. The company first introduced its Flux model for English last October. This new line of models has built-in turn detection and interruption handling.

The company has now introduced a version that supports 10 languages and can handle mid-sentence code switching (people switching from English to Hindi or Spanish). The model focuses entirely on real-time conversations, which is helpful for voice agents.

Deepgram’s CEO, Scott Stephenson, said that Flux’s technical design allows the model to pass the context around better.

“In order to pass the audio Turing test, you have to pass context around. For instance, the end-of-turn detection. That’s context being passed around, noticing that somebody is about to end their sentence or about to switch to a different language when they speak, and then coming back. And this is extra context, which is useful downstream in the system,” he told me.

Stephenson thinks that as Flux is well-suited for real-time voice agents on call, the company’s Net Promoter Score (NPS) will improve. He also thinks that baked-in turn detection is unique to Deepgram’s model and will help the company stand out in the market.

The new Flux model is priced at $0.47 per hour of streaming, as compared to $0.45 per hour of AssemblyAI Universal-3 Pro and $0.96 of Google Cloud Chirp.

(Full interview with Deepgram CEO Scott Stephenson coming this week)

Signals & Experiments

Pronunciation learning has been one of the prime use cases for voice in translation tools. This week, Google Translate rolled out a new feature that lets you practice your pronunciation and enunciation on the tool’s 20th anniversary.

The feature tells you the level of your pronunciation, listens to the ideal pronunciation, and compares both versions as well. At launch, the feature is available in India and the U.S. with support for English, Hindi, and Spanish. This is a great tool for language learning, and I personally believe that hearing and listening to words in a language and then speaking it makes a lot of difference.

In a way, this feature is something out of the Duolingo book, but without structured sessions. I think it would also be useful for Google (or some other tool) to list common words and phrases a traveller can use when they are visiting a particular country. They can have fun while practicing and also preparing for the trip.

Quick Bytes

• Service AI company Avoca, which handles calls for businesses, lets them design campaigns and coaches them on the sales process, reached unicorn valuation with its latest raise. The company said it has raised $125 million across three rounds from investors such as General Catalyst, Meritech Capital Partners, and Kleiner Perkins. But it hasn’t really broken down how much it raised in seed, Series A, and Series B, respectively

• India’s Bajaj Finance handled 52 million calls through AI in FY 2026, the company said during its latest earnings calls. It also said thanks to AI, it is spending a third less on call center calls, as 30% of its outbound is voice agents

• Otter rolled out its enterprise connection suite and a way to search across the entire knowledge base. The company is finally catching up to other players in the market, such as Read AI and Fireflies.ai

• Wispr Flow launched its marketing campaign in the Indian city of Bengaluru on auto rickshaws, but there were mixed reactions as some folks said the language should have been a mix of English and Kannada (the state language). Some others said that the name reminded them of sanitary products. Either way, more people now know about the dictation app

• Taylor Swift joins a horde of celebrities to trademark her voice and protect it from deepfakes. Reuters noted that the pop star’s voice has been used in fake advertisements and political campaigns

• A new survey from Luminate indicates that Gen Z’s interest in AI-generated music has dropped significantly. As we reported last week, despite the rise in the number of AI tracks, Deezer said that overall listenership remains lower than 3%

Partner Spotlight with Atomik Growth

In a chat with a16z, ElevenLabs CEO Mati Staniszewski talks about how he and his co-founder, Piotr Dabkowski, got inspired to solve for badly dubbed movies and created a company that thinks that voice is the new interface for human-computer interaction.

Sponsored content

This newsletter is by Ivan Mehta, a freelance tech reporter at TechCrunch. This newsletter is an attempt at covering what is happening in the industry of voice, audio, and music.

Thank you for tuning in. Keep listening.

Email: voiceaiweek@gmail.com or im@ivanmehta.com

More AI music is being generated, but there are fewer listeners

In the last few years, tools like Suno and Udio have made it very easy for users to create tracks. As a result, music streaming services are seeing an influx of AI-generated music being uploaded every day. Most of them don’t have a clear policy to mark these tracks.

Deezer, however, has made efforts to create tools around identifying fully AI-generated tracks and even weeding out fraudulent tracks. The company released new numbers around AI-generated tracks, and it suggests there is no slowing down. The streaming platform said that 75,000 tracks uploaded on the platform are AI-generated. It noted that this represented roughly 44% of the AI-generated tracks.

In January 2025, Deezer said that 10% of tracks uploaded to the platform were AI-generated. In just over a year, that number has jumped to 44%.

This is only what Deezer’s AI tool has caught onto. By our calculation, the total number of tracks uploaded to the platform has risen from 100,000 per day in January 2025 to around 170,000 per day in April 2026. There could be other heavily AI-aided tracks even if they are not AI-generated.

The rise in AI music is thanks to the rise in AI tooling for the vertical. This week, Suno rose to the #1 spot in the music category in the U.S. According to app analytics company Appfigures, the app has been downloaded over 2.3 million times in the country this year.

Deezer said that while a whole host of music is being uploaded to the platform, only 1-3% of total listens are of AI music. It noted that the purpose of AI tracks is to gain revenue by manufacturing streams. The company noted that up to 85% of the streams generated by fully AI-generated tracks were, in fact, fraudulent in 2025.

A recent study pegged the AI-generated music market to be worth $17.28 billion and would cause artists a revenue loss of over $4.32 billion.

Only a few streaming services are taking a stance against this: Bandcamp has banned AI music, and Deezer is removing these tracks. Platforms like Apple Music and Spotify have made it voluntary to declare if AI was used to create music. We might see more pushback from artists against AI music if money dries up.

Dictation station: new updates from Wispr Flow, Superwhisper, and Nothing

AI-powered transcription apps are finding more users, so there is a race to improve the user interface. Plus, these apps are also trying to use existing phone hardware keys to let users take quick notes.

Wispr Flow released a new insights section for its desktop app. But more importantly, users can now select the level of formatting for transcription from off, light, medium, and high.

The company said that on Android, the Wispr Flow bubble shrinks after five seconds of inactivity, taking up less screen space. Plus, you can also adjust the opacity of the bubble.

This week, hardware company Nothing released a transcription tool called Essential Voice, which supports over 100 languages. Users can be activated through the Essential key or the default keyboard.

There is now a trend of using a hardware key for transcription. Superwhisper released a shortcut that lets you use the iPhone’s action key to activate transcription when the app’s keyboard is open. But if the keyboard is not active, you can use it to take voice notes.

Notably, Wispr Flow also launched a similar feature. We might see more of the “press button to transcribe” features in smartphones in the future.

Signals & Experments

Do you talk to your email? If not, a lot of apps are trying to add features that let you talk to your email. Last week in the newsletter, I covered Y Combinator partner Tom Blomfield’s voicemail app that lets you go through your email through a voice assistant.

Another app that has been doing the rounds is an app called Avec, which is a Tinder-styled Gmail client. The app lets you mark it as done or tackle it later by swiping right or left. The cool thing is that it also has a big button that lets you answer or reply to emails using your voice. The app is using its own transcription engine because it feels that it can provide more personalization and control with its own tech stack.

What’s more, last week, the email app Superhuman also launched a voice reply feature for its users. The company’s CEO, Rahul Vohra, said that, unlike other apps that provide AI-powered transcription, you can also give assistance to the app to write a certain kind of email on your behalf with your voice and your writing style.

I’ve used multiple email clients, and I’ve largely not liked their style of AI writing. At best, I use AI suggestions for a short reply to possibly agree on a meeting time or confirm a detail. If I want to write a longer reply, I usually rely more on voice transcription rather than using AI-powered writing.

With this trend, we could see more and more email apps adopting voice as a reply modality and built-in transcription.

Quick Bytes

• Epic Games is now partnering with Google and ElevenLabs to let developers create voice-powered NPCs in games. Last year, the Screen Actors’ Guild filed a lawsuit against an Epic subsidiary over unfair labor practices.

• Japan is mulling rules to protect celebrity voices against unauthorized AI use. The Japan Times report noted that these voices have been used for sexualized content or for song covers. Actors like Matthew McConaughey have protected their voices under trademark laws.

• A developer is making a local-first transcription app called Steno with Parakeet models, with a very neat interface.

• Truecaller conducted a survey in the U.S., where 53% of the people who have been scammed said that they received a deepfake call. This highlights the rising number of AI calls.

Partner Spotlight with Atomik Growth

In a chat with a16z, ElevenLabs CEO Mati Staniszewski talks about how he and his co-founder, Piotr Dabkowski, got inspired to solve for badly dubbed movies and created a company that thinks that voice is the new interface for human-computer interaction.

Sponsored content

Thank you for tuning in. Keep listening.

Email: voiceaiweek@gmail.com or im@ivanmehta.com

This newsletter is by Ivan Mehta, a consumer tech reporter at TechCrunch. This newsletter is an attempt at covering what is happening in the industry of voice, audio, and music.