Loading video...

Video Failed to Load

Go Home

With Invincible Voice, we help people living with ALS communicate more easily. Encountering Olivier Goy, an entrepreneur who lives with ALS and relentlessly fights to help all patients, made it obvious that our cutting-edge voice AI should help. We turned our Unmute voice-wrapper into a new system that 1/...

kyutai_labs's profile picture

kyutai

27,140 subscribers

11,070 views • 6 months ago •via X (Twitter)

LIVE

Anya Rossi• Live Now

Private livecam show

0 Comments

No comments available

Comments from the original post will appear here

Related Videos

Excited to introduce Fish Speech 1.4 - now open-source and more powerful than ever! 🎉 Our mission is to make cutting-edge voice tech accessible to everyone. What's new: - Trained on 700k hours of multilingual data (up from 200k) - Now supports 8 languages: English, Chinese, German, Japanese, French, Spanish, Korean, and Arabic - Fully open-source, empowering developers and researchers worldwide Key features: - Lightning-fast TTS with ultra-low latency - Instant voice cloning - Self-host or use our cloud service - Simple, flat-rate pricing Try it out: - Playground: - GitHub: - HuggingFace Model: - Demo: - Product Hunt: We can't wait to see what you'll create with Fish Audio. Happy voice building! 🎧🐠
1:23

Excited to introduce Fish Speech 1.4 - now open-source and more powerful than ever! 🎉 Our mission is to make cutting-edge voice tech accessible to everyone. What's new: - Trained on 700k hours of multilingual data (up from 200k) - Now supports 8 languages: English, Chinese, German, Japanese, French, Spanish, Korean, and Arabic - Fully open-source, empowering developers and researchers worldwide Key features: - Lightning-fast TTS with ultra-low latency - Instant voice cloning - Self-host or use our cloud service - Simple, flat-rate pricing Try it out: - Playground: - GitHub: - HuggingFace Model: - Demo: - Product Hunt: We can't wait to see what you'll create with Fish Audio. Happy voice building! 🎧🐠

Fish Audio

149,977 views • 1 year ago

We're launching the ElevenLabs Impact Program to help 1 million people reclaim their voice. We’re partnering with nonprofits to provide free licenses to enhance accessibility, advance education, and improve shared cultural experiences. Our first initiative, made possible through partnership with Bridging Voice and The Scott-Morgan Foundation, is aimed at helping everyone with ALS / MND preserve their voice through free access to our technology to create personalized digital voices that closely match their natural speech. If this can benefit you or a loved one, learn more here: If your nonprofit can benefit from AI Audio, get in touch:
0:58

We're launching the ElevenLabs Impact Program to help 1 million people reclaim their voice. We’re partnering with nonprofits to provide free licenses to enhance accessibility, advance education, and improve shared cultural experiences. Our first initiative, made possible through partnership with Bridging Voice and The Scott-Morgan Foundation, is aimed at helping everyone with ALS / MND preserve their voice through free access to our technology to create personalized digital voices that closely match their natural speech. If this can benefit you or a loved one, learn more here: If your nonprofit can benefit from AI Audio, get in touch:

ElevenLabs

109,635 views • 1 year ago

Excited about the launch of Amazon Nova Sonic, our new speech-to-speech model that helps make AI voice applications feel remarkably natural. It's designed to understand not just what people say, but how they say it – working with tone, style, and conversation flow including pauses and interruptions. Nova Sonic delivers speech understanding and generation through a single, unified model, making it easier for builders to develop voice applications that maintain important context and nuance for customer service, AI agents, and other use cases across industries. It’s available in Amazon Bedrock now. Look forward to seeing what teams build with Nova Sonic!
1:25

Excited about the launch of Amazon Nova Sonic, our new speech-to-speech model that helps make AI voice applications feel remarkably natural. It's designed to understand not just what people say, but how they say it – working with tone, style, and conversation flow including pauses and interruptions. Nova Sonic delivers speech understanding and generation through a single, unified model, making it easier for builders to develop voice applications that maintain important context and nuance for customer service, AI agents, and other use cases across industries. It’s available in Amazon Bedrock now. Look forward to seeing what teams build with Nova Sonic!

Andy Jassy

155,772 views • 1 year ago

Meet MoshiVis🎙️🖼️, the first open-source real-time speech model that can talk about images! It sees, understands, and talks about images — naturally, and out loud. Voice interaction with a compact model endowed with visual understanding opens up new applications, from audio description for the visual impaired to visual access to information. Try it out 👉 Blog post 👉
0:37

Meet MoshiVis🎙️🖼️, the first open-source real-time speech model that can talk about images! It sees, understands, and talks about images — naturally, and out loud. Voice interaction with a compact model endowed with visual understanding opens up new applications, from audio description for the visual impaired to visual access to information. Try it out 👉 Blog post 👉

kyutai

47,949 views • 1 year ago

Today we’re excited to introduce Bland TTS, the first voice AI to cross the uncanny valley. Several months ago, our team solved one-shot style transfer of human speech. That means, from a single, brief MP3, you can clone any voice or remix another clone’s style (tone, cadence, or pronunciations, etc.). With Bland TTS: • Creatives can turn text into realistic sound effects or AI voice tracks with precise control over style and emotion. • Developers can plug in our TTS API to build voice apps fully customized to their liking. • Enterprises can launch AI-driven support lines that sound so natural — customers will save you as a contact (try talking to it on our site). 🧵Check out the thread below or visit our website to get started for free
2:30

Today we’re excited to introduce Bland TTS, the first voice AI to cross the uncanny valley. Several months ago, our team solved one-shot style transfer of human speech. That means, from a single, brief MP3, you can clone any voice or remix another clone’s style (tone, cadence, or pronunciations, etc.). With Bland TTS: • Creatives can turn text into realistic sound effects or AI voice tracks with precise control over style and emotion. • Developers can plug in our TTS API to build voice apps fully customized to their liking. • Enterprises can launch AI-driven support lines that sound so natural — customers will save you as a contact (try talking to it on our site). 🧵Check out the thread below or visit our website to get started for free

Bland

45,967,496 views • 1 year ago

We just shipped something nobody else has. Our STT API doesn't just transcribe, it returns a full voice profile in the same response: emotion, accent, age, pitch, style, each with a confidence score, all in one call. When you feed that into our Realtime API, the LLM decides not just what to say but how to say it. It adds steering tags that our TTS-2 model picks up and uses to guide how it speaks. So, if a user sounds frustrated, the voice responds with empathy. If they're excited, it matches that energy. Here's our very own Clint McLean walking you through how the full pipeline works:
1:24

We just shipped something nobody else has. Our STT API doesn't just transcribe, it returns a full voice profile in the same response: emotion, accent, age, pitch, style, each with a confidence score, all in one call. When you feed that into our Realtime API, the LLM decides not just what to say but how to say it. It adds steering tags that our TTS-2 model picks up and uses to guide how it speaks. So, if a user sounds frustrated, the voice responds with empathy. If they're excited, it matches that energy. Here's our very own Clint McLean walking you through how the full pipeline works:

Inworld AI

250,584 views • 2 months ago

Mistral just released Voxtral TTS, an open-weight voice model that tries to match ElevenLabs while keeping it free. A 3-billionn param model that fits on a laptop and runs 6X faster than real-time speech. It is split into 3 parts that handle different jobs. - The 3.4B part is the main language brain that understands the text and plans what should be spoken. - The 390M part turns that plan into speech features like rhythm, pronunciation, and tone. - The 300M part converts those speech features into the final compressed audio signal you can play. That split helps because each part only does the job it is best at, which can make the whole system faster and lighter than one giant model doing everything alone. Mistral claims about 90ms to first audio, about 6x faster than real-time speech output, about 3GB RAM after quantization, 9 languages, and voice copying from 5s of audio, even across languages. Its internal tests report 62.8% preference over ElevenLabs Flash on default voices and 69.9% on custom voices, but the bigger point is that companies can keep speech data on their own machines. IMO, enterprise buyers usually care most about control, cost, and latency, So this model should give it great push for voice-based AI apps/products.
1:21

Mistral just released Voxtral TTS, an open-weight voice model that tries to match ElevenLabs while keeping it free. A 3-billionn param model that fits on a laptop and runs 6X faster than real-time speech. It is split into 3 parts that handle different jobs. - The 3.4B part is the main language brain that understands the text and plans what should be spoken. - The 390M part turns that plan into speech features like rhythm, pronunciation, and tone. - The 300M part converts those speech features into the final compressed audio signal you can play. That split helps because each part only does the job it is best at, which can make the whole system faster and lighter than one giant model doing everything alone. Mistral claims about 90ms to first audio, about 6x faster than real-time speech output, about 3GB RAM after quantization, 9 languages, and voice copying from 5s of audio, even across languages. Its internal tests report 62.8% preference over ElevenLabs Flash on default voices and 69.9% on custom voices, but the bigger point is that companies can keep speech data on their own machines. IMO, enterprise buyers usually care most about control, cost, and latency, So this model should give it great push for voice-based AI apps/products.

Rohan Paul

15,672 views • 4 months ago

Today we released Meta Spirit LM — our first open source multimodal language model that freely mixes text and speech. Many existing AI voice experiences today use ASR to techniques to process speech before synthesizing with an LLM to generate text — but these approaches compromise the expressive aspects of speech. Using phonetic, pitch and tone tokens, Spirit LM models can overcome these limitations for both inputs and outputs to generate more natural sounding speech while also learning new tasks across ASR, TTS and speech classification. We hope that sharing this work will enable the research community to further new approaches for text and speech integration.
0:36

Today we released Meta Spirit LM — our first open source multimodal language model that freely mixes text and speech. Many existing AI voice experiences today use ASR to techniques to process speech before synthesizing with an LLM to generate text — but these approaches compromise the expressive aspects of speech. Using phonetic, pitch and tone tokens, Spirit LM models can overcome these limitations for both inputs and outputs to generate more natural sounding speech while also learning new tasks across ASR, TTS and speech classification. We hope that sharing this work will enable the research community to further new approaches for text and speech integration.

AI at Meta

351,739 views • 1 year ago

How did a tiny, scrappy team build one of the most powerful AI voice models? In a deep dive with Sesame CTO Ankit Kumar and a16z's Anjney Midha, we explore how Sesame is pushing the boundaries of AI conversation, why it open-sourced its speech generation model, and the power of small teams to outdo much larger AI labs on product focus. A part of their secret: a relentless focus on real-time, natural conversations over raw intelligence, and a deep commitment to voice, personality, and user experience. By opening up its speech generation model, Sesame is paving the way for even more breakthroughs in AI-native conversation 👇
1:04

How did a tiny, scrappy team build one of the most powerful AI voice models? In a deep dive with Sesame CTO Ankit Kumar and a16z's Anjney Midha, we explore how Sesame is pushing the boundaries of AI conversation, why it open-sourced its speech generation model, and the power of small teams to outdo much larger AI labs on product focus. A part of their secret: a relentless focus on real-time, natural conversations over raw intelligence, and a deep commitment to voice, personality, and user experience. By opening up its speech generation model, Sesame is paving the way for even more breakthroughs in AI-native conversation 👇

a16z

29,930 views • 1 year ago

Today, we’re introducing new tools and model updates to help you build, deploy, and scale Voice AI applications.🎙️ 🆕 Speech Understanding: Turn transcripts into actionable data with speaker identification, custom formatting & translation 🆕 LLM Gateway: One API for your voice-to-intelligence pipeline, with GPT, Claude, Gemini & more 🆕 Voice AI Guardrails: End-to-end protection for safe, compliant content Model Upgrades: ✅ 99 languages with auto code-switching ✅ 64% fewer speaker counting errors ✅ 57% better accuracy on critical terms with 1,000-word context Real results: 🔥 Calabrio: 80% boost in customer satisfaction, 22% revenue increase 🔥 Siro: "10/10 customers say 'wow, that insight was crisp'" Build & scale Voice AI in minutes! Try it in our Playground or check the docs. 🚀
0:58

Today, we’re introducing new tools and model updates to help you build, deploy, and scale Voice AI applications.🎙️ 🆕 Speech Understanding: Turn transcripts into actionable data with speaker identification, custom formatting & translation 🆕 LLM Gateway: One API for your voice-to-intelligence pipeline, with GPT, Claude, Gemini & more 🆕 Voice AI Guardrails: End-to-end protection for safe, compliant content Model Upgrades: ✅ 99 languages with auto code-switching ✅ 64% fewer speaker counting errors ✅ 57% better accuracy on critical terms with 1,000-word context Real results: 🔥 Calabrio: 80% boost in customer satisfaction, 22% revenue increase 🔥 Siro: "10/10 customers say 'wow, that insight was crisp'" Build & scale Voice AI in minutes! Try it in our Playground or check the docs. 🚀

AssemblyAI

18,105 views • 9 months ago

Preview of a Gaia 🌱 device! Turn-by-turn voice conversation with your own knowledge assistant (a Gaia node). 💸 $15 worth of hardware (pre-tariff!) 🌱 Voice-to-text, LLM and text-to-voice AI models on Gaia nodes 📖 Open source all-the-way — hardware, firmware, server, AI models 🦀 Rust all-the-way from firmware to AI server Notice how I switched to another language (Chinese) in the middle of the conversation and it seamlessly switched too. Very smart. PS. we are still optimizing performance before the software release.
1:24

Preview of a Gaia 🌱 device! Turn-by-turn voice conversation with your own knowledge assistant (a Gaia node). 💸 $15 worth of hardware (pre-tariff!) 🌱 Voice-to-text, LLM and text-to-voice AI models on Gaia nodes 📖 Open source all-the-way — hardware, firmware, server, AI models 🦀 Rust all-the-way from firmware to AI server Notice how I switched to another language (Chinese) in the middle of the conversation and it seamlessly switched too. Very smart. PS. we are still optimizing performance before the software release.

Michael Yuan

14,983 views • 1 year ago

Learn to build conversational AI voice agents in "Building AI Voice Agents for Production", created in collaboration with LiveKit and RealAvatar, and taught by dsa (Co-founder & CEO of LiveKit), Shayne (Developer Advocate, LiveKit), and Nedelina Teneva (Head of AI at RealAvatar, an AI Fund portfolio company). Voice agents combine speech and reasoning capabilities to enable real-time conversations. They're already being used to support customer service, to improve accessibility in healthcare, for entertainment applications, and for talk therapy. In this course, you’ll learn to build voice agents that listen, reason, and respond naturally. You’ll follow the architecture used to create the "AI Andrew" Avatar, a collaborative project between and RealAvatar that responds to users in what sounds like my voice. You’ll build a voice agent from scratch and deploy it to the cloud, enabling support for many simultaneous users. What you’ll learn: - Understand the fundamentals of voice agents, including key components like speech-to-text (STT), text-to-speech (TTS), and LLMs, and how latency is introduced at each layer. - Explore voice agent architectures and the trade-offs between modular pipelines and speech-to-speech APIs. - Explore how platforms like LiveKit mitigate latency issues with optimized networking infrastructure and low-latency communication protocols. - Learn how to connect client devices to voice agents using WebRTC—and why it outperforms HTTP and WebSocket for low-latency audio streaming. - Incorporate voice activity detection (VAD), end-of-turn detection, and context management to detect turns, handle interruptions, and manage conversational flow. - Understand the trade-offs between latency, quality, and cost in an example in which you build a voice agent and change its voice. - Equip your agent with metrics to measure latency at each stage of the voice pipeline and learn the key levers you can pull to make your agent faster and more responsive. The voice agents built in this course also incorporate voice technology from , a supporting contributor to the project. By the end of this course, you'll have learned the components of an AI voice agent pipeline, combined them into a system with low-latency communication, and deployed them on cloud infrastructure so it scales to many users. I’m looking forward to seeing what voice agents you build from this course! Please sign up here:
2:41

Learn to build conversational AI voice agents in "Building AI Voice Agents for Production", created in collaboration with LiveKit and RealAvatar, and taught by dsa (Co-founder & CEO of LiveKit), Shayne (Developer Advocate, LiveKit), and Nedelina Teneva (Head of AI at RealAvatar, an AI Fund portfolio company). Voice agents combine speech and reasoning capabilities to enable real-time conversations. They're already being used to support customer service, to improve accessibility in healthcare, for entertainment applications, and for talk therapy. In this course, you’ll learn to build voice agents that listen, reason, and respond naturally. You’ll follow the architecture used to create the "AI Andrew" Avatar, a collaborative project between and RealAvatar that responds to users in what sounds like my voice. You’ll build a voice agent from scratch and deploy it to the cloud, enabling support for many simultaneous users. What you’ll learn: - Understand the fundamentals of voice agents, including key components like speech-to-text (STT), text-to-speech (TTS), and LLMs, and how latency is introduced at each layer. - Explore voice agent architectures and the trade-offs between modular pipelines and speech-to-speech APIs. - Explore how platforms like LiveKit mitigate latency issues with optimized networking infrastructure and low-latency communication protocols. - Learn how to connect client devices to voice agents using WebRTC—and why it outperforms HTTP and WebSocket for low-latency audio streaming. - Incorporate voice activity detection (VAD), end-of-turn detection, and context management to detect turns, handle interruptions, and manage conversational flow. - Understand the trade-offs between latency, quality, and cost in an example in which you build a voice agent and change its voice. - Equip your agent with metrics to measure latency at each stage of the voice pipeline and learn the key levers you can pull to make your agent faster and more responsive. The voice agents built in this course also incorporate voice technology from , a supporting contributor to the project. By the end of this course, you'll have learned the components of an AI voice agent pipeline, combined them into a system with low-latency communication, and deployed them on cloud infrastructure so it scales to many users. I’m looking forward to seeing what voice agents you build from this course! Please sign up here:

Andrew Ng

87,484 views • 1 year ago

President Trump has given a voice to so many people who have been affected by the crime in our nation’s capital, and the measures he’s taken are working. It is my hope that he can continue to help other cities get the problem of crime under control.
1:21

President Trump has given a voice to so many people who have been affected by the crime in our nation’s capital, and the measures he’s taken are working. It is my hope that he can continue to help other cities get the problem of crime under control.

Sen. Marsha Blackburn

16,757 views • 10 months ago

NVIDIA just released a new open source transcription model, Nemotron Speech ASR, designed from the ground up for low-latency use cases like voice agents. Here's a voice agent built with this new model. 24ms transcription finalization and total voice-to-voice inference time under 500ms. This agent actually uses *three* NVIDIA open source models: - Nemotron Speech ASR - Nemotron 3 Nano 30GB in a 4-bit quant (released in December) - A preview checkpoint of the upcoming Magpie text-to-speech model These models are all truly open source: weights, training data, training code, and inference code. This is a big deal! Jensen said in the CES keynote yesterday that he expects open source models to catch up to proprietary models this year in a number of categories. NVIDIA is putting their weight behind making this happen. (As Alan Kay said, the best way to predict the future is to invent it.) The code for this agent is open source too, of course. You can deploy it to production with Modal and Pipecat AI cloud, or run locally on an NVIDIA DGX Spark or RTX 5090.
1:24

NVIDIA just released a new open source transcription model, Nemotron Speech ASR, designed from the ground up for low-latency use cases like voice agents. Here's a voice agent built with this new model. 24ms transcription finalization and total voice-to-voice inference time under 500ms. This agent actually uses *three* NVIDIA open source models: - Nemotron Speech ASR - Nemotron 3 Nano 30GB in a 4-bit quant (released in December) - A preview checkpoint of the upcoming Magpie text-to-speech model These models are all truly open source: weights, training data, training code, and inference code. This is a big deal! Jensen said in the CES keynote yesterday that he expects open source models to catch up to proprietary models this year in a number of categories. NVIDIA is putting their weight behind making this happen. (As Alan Kay said, the best way to predict the future is to invent it.) The code for this agent is open source too, of course. You can deploy it to production with Modal and Pipecat AI cloud, or run locally on an NVIDIA DGX Spark or RTX 5090.

kwindla

274,474 views • 6 months ago

Mati Staniszewski on why ElevenLabs is betting on a cascaded approach for voice agents: “Our approach, as you think about voice agents, conversational agents, is effectively a cascaded approach. You use transcription or speech-to-text, an LLM, text-to-speech, and orchestrate all of that together. Then you have speech-to-speech, which goes directly from speech, and there’s a speech response on the other side. Today, we are optimizing heavily on a cascaded approach. As we work with a lot of the businesses and enterprises, they will need that visibility into what happens. They will want to execute certain tasks on top of that. They want good visibility into each of the steps and great accuracy of all the models. But beyond that, they can abstract away what’s the LLM layer, what’s the intelligence layer, and the integrations are easier in that system. That’s where we are betting a lot of the research work on how you can make that great, and we think we can make that great.” John Collison Mati Staniszewski ElevenLabs
1:09

Mati Staniszewski on why ElevenLabs is betting on a cascaded approach for voice agents: “Our approach, as you think about voice agents, conversational agents, is effectively a cascaded approach. You use transcription or speech-to-text, an LLM, text-to-speech, and orchestrate all of that together. Then you have speech-to-speech, which goes directly from speech, and there’s a speech response on the other side. Today, we are optimizing heavily on a cascaded approach. As we work with a lot of the businesses and enterprises, they will need that visibility into what happens. They will want to execute certain tasks on top of that. They want good visibility into each of the steps and great accuracy of all the models. But beyond that, they can abstract away what’s the LLM layer, what’s the intelligence layer, and the integrations are easier in that system. That’s where we are betting a lot of the research work on how you can make that great, and we think we can make that great.” John Collison Mati Staniszewski ElevenLabs

Stripe

17,118 views • 3 months ago

Voice In My Head, a collaboration with Kyle McDonald. With the proliferation of generated content, AI now seeps constantly into our consciousness. What happens when it begins to intervene directly into your thoughts? Where the people you interact with, the things you do, are guided by an AI enhanced voice that speaks to you the way you’d like to be spoken to. Voice In My Head explores the implications for AI (ChatGPT) to listen and intervene in your social experience in real-time, augmenting your personality. The piece begins with an onboarding session where you place a bud in your ear and the voice asks you to reflect on the inner voice you were born with. What if it could be more caring? Less obsessive? Less judgmental? More helpful? What if you could change your inner monologue? As you respond to the onboarding questions, it clones the sound of your voice and uses it to speak to you. Then you go out into the world, as the voice follows along and offers commentary and direction. The resulting performance calls into question how natural vs synthetic each person’s thoughts actually are. Do any of us have our own point of view? Trailer directed by David Leonard.
1:19

Voice In My Head, a collaboration with Kyle McDonald. With the proliferation of generated content, AI now seeps constantly into our consciousness. What happens when it begins to intervene directly into your thoughts? Where the people you interact with, the things you do, are guided by an AI enhanced voice that speaks to you the way you’d like to be spoken to. Voice In My Head explores the implications for AI (ChatGPT) to listen and intervene in your social experience in real-time, augmenting your personality. The piece begins with an onboarding session where you place a bud in your ear and the voice asks you to reflect on the inner voice you were born with. What if it could be more caring? Less obsessive? Less judgmental? More helpful? What if you could change your inner monologue? As you respond to the onboarding questions, it clones the sound of your voice and uses it to speak to you. Then you go out into the world, as the voice follows along and offers commentary and direction. The resulting performance calls into question how natural vs synthetic each person’s thoughts actually are. Do any of us have our own point of view? Trailer directed by David Leonard.

Lauren Lee McCarthy

96,073 views • 2 years ago

Cerebras inference is very fast. So fast that it changes how we think about configuring our LLMs for voice agent use cases. Kimi K2.6 is a 1T parameter reasoning model that Cerebras serves at 650 - 1,000 tokens per second (end-to-end throughput), with time to first token metrics as low as 150ms (latency). These numbers are two to three times faster than other similarly capable models. The biggest lever we get from this kind of speed is that we can use the model in reasoning mode, and still have excellent "time to first non-thinking token." This solves a big pain point we have in 2026 for voice agent use cases. Almost all recent innovation in post-training has focused on making models good at reasoning ("test time compute"). This is great, but it makes the user-facing model latency much, much slower. Which is a problem for conversational voice agents. We can run Kimi K2.6 with reasoning turned on, and get responses faster than other models produce with reasoning disabled. On my 30-turn voice agent benchmark, Kimi K2.6 with reasoning enabled ties GPT 5.1 and Haiku 4.5 with reasoning disabled, and is still about 200ms seconds faster! On my primary task agent benchmark, Kimi K2.6 is now the #2 model. It ranks just behind Gemini 3.5 Flash in "high" reasoning mode, and tied with GLM 5, Sonnet 4.6, and GPT 5.4 with reasoning set to "low." But Kimi K2.6 completes each turn in the agent loop in under 500ms. The other four models are all at least 3x slower. (Models only qualify for this benchmark if they can complete task turns at a P50 <4s.) A couple of other things that this speed buys us, for production voice agents: - Tool calls happen fast enough that we don't have to work around tool call latency in our pipeline design. - We can prompt the model to output structured data at the beginning of a response, followed by plain text for voice generation. This opens up possibilities like asking the model to do complex classification/generation tasks that influence the rest of the pipeline. For example, the model could create a detailed style prompt for a steerable TTS model, for each individual conversation turn. And, of course, you can use Kimi K2.6 with reasoning turned off. Cerebras calls this "instant" mode. Here's a video of a Cerebras Kimi K2.6 voice agent with voice-to-voice response time, measured at the client, under 500ms. This is the true response latency as perceived by the user, including all network and audio codec overhead, transcription and turn detection, Kimi K2.6 token generation, and voice generation. 500ms is, effectively, instant. So the Cerebras naming for this mode is a propos. :-)
0:40

Cerebras inference is very fast. So fast that it changes how we think about configuring our LLMs for voice agent use cases. Kimi K2.6 is a 1T parameter reasoning model that Cerebras serves at 650 - 1,000 tokens per second (end-to-end throughput), with time to first token metrics as low as 150ms (latency). These numbers are two to three times faster than other similarly capable models. The biggest lever we get from this kind of speed is that we can use the model in reasoning mode, and still have excellent "time to first non-thinking token." This solves a big pain point we have in 2026 for voice agent use cases. Almost all recent innovation in post-training has focused on making models good at reasoning ("test time compute"). This is great, but it makes the user-facing model latency much, much slower. Which is a problem for conversational voice agents. We can run Kimi K2.6 with reasoning turned on, and get responses faster than other models produce with reasoning disabled. On my 30-turn voice agent benchmark, Kimi K2.6 with reasoning enabled ties GPT 5.1 and Haiku 4.5 with reasoning disabled, and is still about 200ms seconds faster! On my primary task agent benchmark, Kimi K2.6 is now the #2 model. It ranks just behind Gemini 3.5 Flash in "high" reasoning mode, and tied with GLM 5, Sonnet 4.6, and GPT 5.4 with reasoning set to "low." But Kimi K2.6 completes each turn in the agent loop in under 500ms. The other four models are all at least 3x slower. (Models only qualify for this benchmark if they can complete task turns at a P50 <4s.) A couple of other things that this speed buys us, for production voice agents: - Tool calls happen fast enough that we don't have to work around tool call latency in our pipeline design. - We can prompt the model to output structured data at the beginning of a response, followed by plain text for voice generation. This opens up possibilities like asking the model to do complex classification/generation tasks that influence the rest of the pipeline. For example, the model could create a detailed style prompt for a steerable TTS model, for each individual conversation turn. And, of course, you can use Kimi K2.6 with reasoning turned off. Cerebras calls this "instant" mode. Here's a video of a Cerebras Kimi K2.6 voice agent with voice-to-voice response time, measured at the client, under 500ms. This is the true response latency as perceived by the user, including all network and audio codec overhead, transcription and turn detection, Kimi K2.6 token generation, and voice generation. 500ms is, effectively, instant. So the Cerebras naming for this mode is a propos. :-)

kwindla

40,319 views • 2 months ago

Taking talking shop to a whole new level. We just shipped Glean’s real-time voice capability, powered by OpenAI’s newest speech model GPT-Realtime-2. Grounded in the context across your org, it feels like a real AI coworker and can keep up with how work gets finished. In internal evals, GPT-Realtime-2 delivered a 42.9% relative increase in helpfulness over its previous version. Give it a try. It speaks for itself. OpenAIDevs
1:55

Taking talking shop to a whole new level. We just shipped Glean’s real-time voice capability, powered by OpenAI’s newest speech model GPT-Realtime-2. Grounded in the context across your org, it feels like a real AI coworker and can keep up with how work gets finished. In internal evals, GPT-Realtime-2 delivered a 42.9% relative increase in helpfulness over its previous version. Give it a try. It speaks for itself. OpenAIDevs

Glean

16,588 views • 2 months ago

Walmart updated their new AI call system for when customer call in It now says, “This call and your voice may be recorded for quality or other business purposes” You won’t believe what their “other business serves” are… It is not limited to AI training, voice cloning datasets, emotion tracking, behavioral analysis, customer profiling, speech recognition systems So when you call into Walmart they are literally recording. then storing your voice. It can possibly be used for things like “voice cloning” and more This is horrifying and it’s really happening Here are some of the more controversial uses: - behavioral profiling - emotion detection - cross-linking voice data with customer profiles - long-term biometric storage - AI model training without clear opt-in transparency (You are literally just calling in to talk to customer service) Even worse, modern AI systems can extract a lot more information from speech than people realize: - accent - emotional state - stress - probable age range - speaking habits - identity likelihood - purchase intent - deception indicators (though these systems are often unreliable) All this and voice cloning…. Some states have started to pass legislation to stop this but it’s still in the early stages so companies are getting away with it
0:40

Walmart updated their new AI call system for when customer call in It now says, “This call and your voice may be recorded for quality or other business purposes” You won’t believe what their “other business serves” are… It is not limited to AI training, voice cloning datasets, emotion tracking, behavioral analysis, customer profiling, speech recognition systems So when you call into Walmart they are literally recording. then storing your voice. It can possibly be used for things like “voice cloning” and more This is horrifying and it’s really happening Here are some of the more controversial uses: - behavioral profiling - emotion detection - cross-linking voice data with customer profiles - long-term biometric storage - AI model training without clear opt-in transparency (You are literally just calling in to talk to customer service) Even worse, modern AI systems can extract a lot more information from speech than people realize: - accent - emotional state - stress - probable age range - speaking habits - identity likelihood - purchase intent - deception indicators (though these systems are often unreliable) All this and voice cloning…. Some states have started to pass legislation to stop this but it’s still in the early stages so companies are getting away with it

Wall Street Apes

113,181 views • 2 months ago