Video yükleniyor...

Video Yüklenemedi

Bu video yüklenirken bir sorun oluştu. Bu geçici bir ağ sorunundan kaynaklanıyor olabilir veya video kullanılamıyor olabilir.

Ana Sayfaya Dön

Introducing Jarvis Bench v0.5, Voice Arena's conversational agent benchmark. We've been obsessed with one question at VoiceArena: why do voice agent demos sound incredible, benchmarks say models are near-perfect, and yet you probably didn't have a single real conversation with a voice agent in the last 24 hours? Here's... show more

Shobhit Banga

3,143 subscribers

568,930 görüntüleme • 10 gün önce •via X (Twitter)

Eğitim Bilim & Teknoloji

Anya Rossi• Live Now

Private livecam show

12 Yorum

Shobhit Banga profil fotoğrafı

Shobhit Banga10 gün önce

The first board is Naturalness: how much each call sounds like a real human conversation. The human agent sits at 1244 Elo. Every model is a wide gap below the human: › gpt-realtime 984 (-260) › Gemini 3.1 Flash Live 970 (-274) › GPT Live 1 967 (-277) › Grok Voice 920 (-324) › Inworld cascaded 915 (-329) The best model wins just 18% of blind calls against a human. No model is above 18%. Sounding human is where the gap is still widest.

Shobhit Banga profil fotoğrafı

Shobhit Banga10 gün önce

Read every pair head to head and the same thing holds. The human agent is favoured against every single model: 82% against gpt-realtime, and up to 87% against Grok Voice and Inworld. The models sit within 60-40 of each other, so the race between them is close. The race against a person is not: none has crossed 50% yet.

Shobhit Banga profil fotoğrafı

Shobhit Banga10 gün önce

Here's one of those calls: gpt-realtime, our top model on Humanness, next to a real caller, on a gift-budget question. Same conversation, side by side. Play it and see how far 18% actually sounds.

Shobhit Banga profil fotoğrafı

Shobhit Banga10 gün önce

On Task Completion, though, it's a different story, the gap is much closer. Votes are still coming in and we will keep updating the leaderboard. This is a work in progress. Tag the model labs you would like to see added to this benchmark.

Rohan Paul profil fotoğrafı

Rohan Paul10 gün önce

@voicearena_ai yes, I suspect voice benchmarks are going to become much more multidimensional than text benchmarks.

AshutoshShrivastava profil fotoğrafı

AshutoshShrivastava10 gün önce

@voicearena_ai most systems are trained to sound like polished audiobook narrators. real human dialogue is messy, with false starts and micro-interruptions. until models learn acoustic imperfection, that naturalness gap stays wide open

Megha Vishwanath profil fotoğrafı

Megha Vishwanath10 gün önce

@voicearena_ai This is great and much needed! Congratulations to the entire team 🙌🏼

Spandan Singh profil fotoğrafı

Spandan Singh10 gün önce

@voicearena_ai Demos and clean benchmarks fail the same way: they never see barge-in on dirty audio mid-tool-call. Blind human ratings on real conversations is the right bar. Task completion without sounding robotic is the harder half.

Aaliya profil fotoğrafı

Aaliya10 gün önce

@voicearena_ai I think real conversations are a much better test than demos.

Anissa profil fotoğrafı

Anissa10 gün önce

@voicearena_ai Real conversations reveal what benchmarks miss

Khawaja Ahmed Roshan profil fotoğrafı

Khawaja Ahmed Roshan10 gün önce

@voicearena_ai blind votes are closer to the real test than a transcript score. the failure mode is usually turn-taking and recovery after a bad ASR guess, not the first answer.

NAMAN RAJ profil fotoğrafı

NAMAN RAJ10 gün önce

@voicearena_ai @shobhitbanga sounds like your 'bench' has some ghostly AI 'users' 🤖😅

Benzer Videolar

Introducing Expressive Mode for ElevenAgents - voice agents so expressive, they blur the line between AI and human conversations. This is an unedited recording of an agent empathizing with a customer at peak frustration.

Introducing Expressive Mode for ElevenAgents - voice agents so expressive, they blur the line between AI and human conversations. This is an unedited recording of an agent empathizing with a customer at peak frustration.

ElevenLabs

11,057,621 görüntüleme • 7 ay önce

Introducing PhoneLLM, an open model for voice agents. GPT 5.6 Terra performance on typical voice agent tasks at 1/3 the latency and 1/18 the cost. For voice agents, we need models that are both very low latency and very good at tool calling and instruction following. There's a trade-off here, and we often have to compromise on either latency or capability when building voice agents. With PhoneLLM (and the training and data stack that made this model possible) we're fixing this problem. For the last couple of years, most of the effort in frontier model development has gone towards leveraging test-time compute. Which is awesome! Models of all shapes and sizes are available that perform really, really well ... if you have "thinking" turned on for your model. But if you need your agent to respond at voice conversation speed, you can't use thinking models. PhoneLLM is a full-weights fine-tune of NVIDIA Nemotron Nano 30B. We trained on a wide range of real-world telephone and customer support use cases. The training focused on taking the excellent Nano 30B base capabilities and teaching the model to do typical voice agent tasks with thinking disabled. The results are really good: accurate tool calling and concise, on-topic responses in long conversations. And fast: TTFAT measured server-side is <100ms if you run PhoneLLM on a lightly loaded B200. :-) But seriously, when we characterize model latency, we do it with full, end-to-end, batched request simulations using real Pipecat voice agent pipelines. You can serve more than 80 concurrent agents on a single B200 with P95 end-to-end TTFAT <600ms. Including network overhead. That's an LLM cost-per-minute around $0.0025. (1/4 of a cent.) At a latency lower than any third-party API offers today. More details about this model, including weights on Hugging Face, how to spin it up with one click on Modal, and a starter project repo you can clone, are in the thread ...

Introducing PhoneLLM, an open model for voice agents. GPT 5.6 Terra performance on typical voice agent tasks at 1/3 the latency and 1/18 the cost. For voice agents, we need models that are both very low latency and very good at tool calling and instruction following. There's a trade-off here, and we often have to compromise on either latency or capability when building voice agents. With PhoneLLM (and the training and data stack that made this model possible) we're fixing this problem. For the last couple of years, most of the effort in frontier model development has gone towards leveraging test-time compute. Which is awesome! Models of all shapes and sizes are available that perform really, really well ... if you have "thinking" turned on for your model. But if you need your agent to respond at voice conversation speed, you can't use thinking models. PhoneLLM is a full-weights fine-tune of NVIDIA Nemotron Nano 30B. We trained on a wide range of real-world telephone and customer support use cases. The training focused on taking the excellent Nano 30B base capabilities and teaching the model to do typical voice agent tasks with thinking disabled. The results are really good: accurate tool calling and concise, on-topic responses in long conversations. And fast: TTFAT measured server-side is <100ms if you run PhoneLLM on a lightly loaded B200. :-) But seriously, when we characterize model latency, we do it with full, end-to-end, batched request simulations using real Pipecat voice agent pipelines. You can serve more than 80 concurrent agents on a single B200 with P95 end-to-end TTFAT <600ms. Including network overhead. That's an LLM cost-per-minute around $0.0025. (1/4 of a cent.) At a latency lower than any third-party API offers today. More details about this model, including weights on Hugging Face, how to spin it up with one click on Modal, and a starter project repo you can clone, are in the thread ...

kwindla

331,598 görüntüleme • 29 gün önce

Today, we’re making Voice Agents available to everyone. With Voice Agents, you can build and scale intelligent, human-like agents that understand context, remember past conversations, and get better at achieving your business outcomes with every interaction. Our Voice Agents have already powered more than 350 million conversations across enterprise deployments. Operating at this scale has helped us refine the platform, deepen its capabilities, and build out the complete stack needed to handle the full complexity of voice interactions. The platform brings together the tools and systems to personalise, orchestrate, govern, and launch Voice Agents at scale. Start building your Voice Agent today:

Today, we’re making Voice Agents available to everyone. With Voice Agents, you can build and scale intelligent, human-like agents that understand context, remember past conversations, and get better at achieving your business outcomes with every interaction. Our Voice Agents have already powered more than 350 million conversations across enterprise deployments. Operating at this scale has helped us refine the platform, deepen its capabilities, and build out the complete stack needed to handle the full complexity of voice interactions. The platform brings together the tools and systems to personalise, orchestrate, govern, and launch Voice Agents at scale. Start building your Voice Agent today:

Sarvam

280,019 görüntüleme • 1 ay önce

🚨NEW: CHATGPT LIVE VOICE MODEL NOW DELIVERS REAL-TIME CONVERSATIONAL TRANSLATION Built on a full-duplex architecture, GPT-Live is a new generation of voice models that can listen and speak simultaneously, enabling more natural human-AI interactions. The breakthrough unlocks fluid back-and-forth conversations, real-time translation, and a far more human sense of timing and context.

🚨NEW: CHATGPT LIVE VOICE MODEL NOW DELIVERS REAL-TIME CONVERSATIONAL TRANSLATION Built on a full-duplex architecture, GPT-Live is a new generation of voice models that can listen and speak simultaneously, enabling more natural human-AI interactions. The breakthrough unlocks fluid back-and-forth conversations, real-time translation, and a far more human sense of timing and context.

Coin Bureau

29,085 görüntüleme • 2 ay önce

Sound on for this one. We wired up OpenAI's new STT and TTS models into a single voice agent. The results are super fun! Link to a live demo in the next tweet.

Sound on for this one. We wired up OpenAI's new STT and TTS models into a single voice agent. The results are super fun! Link to a live demo in the next tweet.

dsa

32,310 görüntüleme • 1 yıl önce

Introducing our unified Voice Agent API! ⚡️ Quickly and easily build voice agents that listen, think, and respond in real-time, as naturally as a human can. Powered by the industry's fastest, most performant speech recognition and voice synthesis models to deliver: -Natural-sounding real-time conversations -Revolutionary interruption handling -Choice of open source, closed source, or BYO LLM -Low cost See our drive-thru agent in action, smoothly handling interruptions and complex orders in the noisy streets of SF. Learn how to get access (🧵)

Introducing our unified Voice Agent API! ⚡️ Quickly and easily build voice agents that listen, think, and respond in real-time, as naturally as a human can. Powered by the industry's fastest, most performant speech recognition and voice synthesis models to deliver: -Natural-sounding real-time conversations -Revolutionary interruption handling -Choice of open source, closed source, or BYO LLM -Low cost See our drive-thru agent in action, smoothly handling interruptions and complex orders in the noisy streets of SF. Learn how to get access (🧵)

Deepgram

726,214 görüntüleme • 2 yıl önce

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

87,868 görüntüleme • 1 yıl önce

Today we’re launching our first homegrown AI model: an open source turn detection model for building voice agents. Instead of relying solely on voice activity detection (VAD), which only considers when a user is speaking, our model also considers what has and is being said in the context of a conversation and predicts when a user is finished expressing their thoughts before the agent responds. Conversations with AI voice agents using this new model flow much more naturally without constant interruptions from the AI— check it out (more videos, details, and code in the thread):

Today we’re launching our first homegrown AI model: an open source turn detection model for building voice agents. Instead of relying solely on voice activity detection (VAD), which only considers when a user is speaking, our model also considers what has and is being said in the context of a conversation and predicts when a user is finished expressing their thoughts before the agent responds. Conversations with AI voice agents using this new model flow much more naturally without constant interruptions from the AI— check it out (more videos, details, and code in the thread):

LiveKit

126,983 görüntüleme • 1 yıl önce

Introducing Diarization Bench, Voice Arena's benchmark for who spoke when. The metric we measure is diarization error rate, or DER. It measures how much of a conversation a model attaches to the wrong person. Every second it misses, invents, or gives to another speaker counts against it. Here's what we did. 139 conversations recorded for this board, in rooms and on calls, with people interrupting each other throughout. 12 systems scored under a single goal: testing models in the real-world conditions they are actually deployed in. NVIDIA's Nemotron 3 leads at 14.72%. 🧵

Introducing Diarization Bench, Voice Arena's benchmark for who spoke when. The metric we measure is diarization error rate, or DER. It measures how much of a conversation a model attaches to the wrong person. Every second it misses, invents, or gives to another speaker counts against it. Here's what we did. 139 conversations recorded for this board, in rooms and on calls, with people interrupting each other throughout. 12 systems scored under a single goal: testing models in the real-world conditions they are actually deployed in. NVIDIA's Nemotron 3 leads at 14.72%. 🧵

Shobhit Banga

629,621 görüntüleme • 2 gün önce

Agent-native products are coming. Every product on the internet was built for a human with eyes, a cursor, and a credit card. Agents have none of those things. Most companies are teaching agents to pretend to be humans. That's a hack. The real opportunity is products designed for agents from scratch. Everything inverts: • Discovery → protocol registries, not ads and billboards • Trust → machine-readable reputation, not brand • Onboarding → full capabilities upfront, not a narrow slice • Payments → spend authorization, not checkout flows • Retention → zero. Agents switch between API calls. 30 years of human product design. Day one of agent product design.

Fareed Mosavat

25,837 görüntüleme • 4 ay önce

Anthropic's multi-agent researcher just explained how to actually build a graph of Claude agents: "Multi-agent is where fundamentally you have multiple agents or multiple Claudes working at the same time - one parent agent delegates tasks to five sub-agents that can each then work in parallel." the real architecture: parent agent → delegates to sub-agents (as tools) → sub-agents run in parallel → results roll back up the failure mode nobody talks about: "Overbuilt multi-agent systems spend too much time just talking back and forth with each other and not actually making progress on the main task." sound familiar? that's every bloated human org chart too in this 19-minute conversation, an Anthropic researcher lays out the real framework: parent → sub-agents → tools → context protection = the actual shape of a working agent graph

Anthropic's multi-agent researcher just explained how to actually build a graph of Claude agents: "Multi-agent is where fundamentally you have multiple agents or multiple Claudes working at the same time - one parent agent delegates tasks to five sub-agents that can each then work in parallel." the real architecture: parent agent → delegates to sub-agents (as tools) → sub-agents run in parallel → results roll back up the failure mode nobody talks about: "Overbuilt multi-agent systems spend too much time just talking back and forth with each other and not actually making progress on the main task." sound familiar? that's every bloated human org chart too in this 19-minute conversation, an Anthropic researcher lays out the real framework: parent → sub-agents → tools → context protection = the actual shape of a working agent graph

rewind

27,195 görüntüleme • 2 ay önce

Introducing LobeHub: Agent teammates that grow with you. LobeHub is the ultimate space for work and life: to find, build, and collaborate with agent teammates that grow with you. We’re building the world’s first and largest human–agent co-evolving network. Two years ago, we built LobeChat, an open-source interface for using different AI models. Today, LobeChat has 70k+ GitHub stars and serves 6M+ users worldwide. How to fully unlock the power of models has always been a shared mission between us and the community. We started with interaction — a fundamentally new, agent-first experience. Agents are no longer passive tools invoked in a single conversation. They should be proactive, always-on units of work. Treating agents as the minimal atomic unit is also the core of our agent harness infra. Today’s agents are mostly one-off executors. Even with memory, it’s often global — and hallucinates. We build long-term agent teammates that evolve with users. Each agent has its own dedicated memory space, editable by users, allowing humans and agents to co-evolve over time. This, in turn, allows us to design clearer rewards for reinforcement learning and create cleaner environments for continual learning. Agent teammates can work in groups. Through a multi-agent system, agent groups operate faster, more cost-effective, and go beyond what single-agent systems can achieve. For example, a single agent often requires heavy user involvement to proceed step by step, whereas LobeHub can execute the same work from a single instruction, with a supervisor orchestrating agents that run in parallel or debate to produce better results. We are building the collaboration network among agent teammates — and between humans and agent teammates as well. Ease of use matters. AI intelligence and shared human intelligence are equally important. With simple instructions and tool selection, you can effortlessly build and team up with agent coworkers to deliver complex, systematic work — even assembling a quant team to execute trades. Through the LobeHub community, anyone can discover, reuse, and remix agents and agent groups, customizing them to fit their own workflows, preferences, and needs. Last but not least, our vision started with LobeChat: multi-model support is the most efficient approach for users. We believe different models excel in different scenarios. By routing across multiple models, LobeHub improves cost efficiency and unlocks capabilities that a single-model setup cannot easily support.

Introducing LobeHub: Agent teammates that grow with you. LobeHub is the ultimate space for work and life: to find, build, and collaborate with agent teammates that grow with you. We’re building the world’s first and largest human–agent co-evolving network. Two years ago, we built LobeChat, an open-source interface for using different AI models. Today, LobeChat has 70k+ GitHub stars and serves 6M+ users worldwide. How to fully unlock the power of models has always been a shared mission between us and the community. We started with interaction — a fundamentally new, agent-first experience. Agents are no longer passive tools invoked in a single conversation. They should be proactive, always-on units of work. Treating agents as the minimal atomic unit is also the core of our agent harness infra. Today’s agents are mostly one-off executors. Even with memory, it’s often global — and hallucinates. We build long-term agent teammates that evolve with users. Each agent has its own dedicated memory space, editable by users, allowing humans and agents to co-evolve over time. This, in turn, allows us to design clearer rewards for reinforcement learning and create cleaner environments for continual learning. Agent teammates can work in groups. Through a multi-agent system, agent groups operate faster, more cost-effective, and go beyond what single-agent systems can achieve. For example, a single agent often requires heavy user involvement to proceed step by step, whereas LobeHub can execute the same work from a single instruction, with a supervisor orchestrating agents that run in parallel or debate to produce better results. We are building the collaboration network among agent teammates — and between humans and agent teammates as well. Ease of use matters. AI intelligence and shared human intelligence are equally important. With simple instructions and tool selection, you can effortlessly build and team up with agent coworkers to deliver complex, systematic work — even assembling a quant team to execute trades. Through the LobeHub community, anyone can discover, reuse, and remix agents and agent groups, customizing them to fit their own workflows, preferences, and needs. Last but not least, our vision started with LobeChat: multi-model support is the most efficient approach for users. We believe different models excel in different scenarios. By routing across multiple models, LobeHub improves cost efficiency and unlocks capabilities that a single-model setup cannot easily support.

LobeHub

185,401 görüntüleme • 8 ay önce

Async, automatic, non-blocking context compaction for long-running agents. Last week I gave a talk called Space Machine Sandboxes at the Daytona AI builders meetup about patterns for long-running agents. I work a lot on voice AI agents, which are fundamentally multi-turn, long-context loops. I also build lots of other AI agent stuff, often as part of bigger systems that include voice. One of the patterns I showed in the talk is non-blocking compaction. Here's a short clip.

Async, automatic, non-blocking context compaction for long-running agents. Last week I gave a talk called Space Machine Sandboxes at the Daytona AI builders meetup about patterns for long-running agents. I work a lot on voice AI agents, which are fundamentally multi-turn, long-context loops. I also build lots of other AI agent stuff, often as part of bigger systems that include voice. One of the patterns I showed in the talk is non-blocking compaction. Here's a short clip.

kwindla

25,877 görüntüleme • 8 ay önce

Bluejay (Bluejay) is the E2E testing platform for voice AI agents. On Bluejay, simulate a month's worth of customer interaction with your agent in just 5 minutes. Congrats on the launch, @faraz_ml and rohan!

Bluejay (Bluejay) is the E2E testing platform for voice AI agents. On Bluejay, simulate a month's worth of customer interaction with your agent in just 5 minutes. Congrats on the launch, @faraz_ml and rohan!

Y Combinator

24,582 görüntüleme • 1 yıl önce

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. :-)

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

40,593 görüntüleme • 4 ay önce

New course: Add voice to your AI agents and applications, built with Vocal Bridge (disclosure: an AI Fund portfolio company) and taught by its CEO Ashwyn Sharma. Voice applications historically required making a hard tradeoff: using fast voice-to-voice models that sacrifice reliability, or accurate speech-to-text pipelines that add latency. This course teaches you how to build voice agents that are both reliable and fast. You'll build three types of voice-enabled applications: a voice-interactive game where voice commands and mouse clicks work together over a single channel, an agent that gains a voice in about 10 lines of code without touching its prompts or tools, and an agent that places outbound phone calls using a make_phone_call function. Skills you'll gain: - Add a voice layer to an existing agent without rewriting your prompts, RAG pipeline, or tools - Give an agent the ability to place outbound calls and stream transcripts back live - Set up voice evaluation to score calls, catch regressions, and improve quality before deployment Join and add voice to your agents without overhauling your architecture:

New course: Add voice to your AI agents and applications, built with Vocal Bridge (disclosure: an AI Fund portfolio company) and taught by its CEO Ashwyn Sharma. Voice applications historically required making a hard tradeoff: using fast voice-to-voice models that sacrifice reliability, or accurate speech-to-text pipelines that add latency. This course teaches you how to build voice agents that are both reliable and fast. You'll build three types of voice-enabled applications: a voice-interactive game where voice commands and mouse clicks work together over a single channel, an agent that gains a voice in about 10 lines of code without touching its prompts or tools, and an agent that places outbound phone calls using a make_phone_call function. Skills you'll gain: - Add a voice layer to an existing agent without rewriting your prompts, RAG pipeline, or tools - Give an agent the ability to place outbound calls and stream transcripts back live - Set up voice evaluation to score calls, catch regressions, and improve quality before deployment Join and add voice to your agents without overhauling your architecture:

Andrew Ng

96,370 görüntüleme • 3 ay önce

Launching 'Expert Agents' on Agent Two. A specialised creative crew of AI agents, trained on real creative expertise. Name the role and it's filled: creative director, DOP, storyboard artist, and more. And they talk to each other. Independently. In parallel. You're no longer briefing each agent one by one: you're running a crew, exactly like on set. Expert agents. Now in conversation. Agent Two is frontier intelligence for creative work. 4x faster, at half the price.

Launching 'Expert Agents' on Agent Two. A specialised creative crew of AI agents, trained on real creative expertise. Name the role and it's filled: creative director, DOP, storyboard artist, and more. And they talk to each other. Independently. In parallel. You're no longer briefing each agent one by one: you're running a crew, exactly like on set. Expert agents. Now in conversation. Agent Two is frontier intelligence for creative work. 4x faster, at half the price.

Invideo

3,411,230 görüntüleme • 1 ay önce

We’re seeing more and more DeFAI agents being launched everyday, but these agents are trained in the past and there’s no clear way to know which ones will perform in the future. That’s why we're launching ALFA with NEAR Protocol. ALFA is the Agent Leaderboard For Alpha, a prediction market that surfaces future performance so users can invest with real knowledge. Let’s introduce the first batch of agents:

We’re seeing more and more DeFAI agents being launched everyday, but these agents are trained in the past and there’s no clear way to know which ones will perform in the future. That’s why we're launching ALFA with NEAR Protocol. ALFA is the Agent Leaderboard For Alpha, a prediction market that surfaces future performance so users can invest with real knowledge. Let’s introduce the first batch of agents:

Fraction AI

28,474 görüntüleme • 1 yıl önce

Introducing October, the world's most advanced agent orchestrator Claude? Codex? Grok? Cursor? Which model? Which harness? How much quota is left? You shouldn't have to care Give October a task. It builds the team, connects the agents with native agent agent comms, and orchestrates the work based on the task, your taste, budget and remaining usage (or build the team yourself) And it's multiplayer: your team's agents can work with yours, live on the same canvas

Introducing October, the world's most advanced agent orchestrator Claude? Codex? Grok? Cursor? Which model? Which harness? How much quota is left? You shouldn't have to care Give October a task. It builds the team, connects the agents with native agent agent comms, and orchestrates the work based on the task, your taste, budget and remaining usage (or build the team yourself) And it's multiplayer: your team's agents can work with yours, live on the same canvas

Harsh Savergaonkar

22,604 görüntüleme • 8 gün önce