Loading video...

Video Failed to Load

There was a problem loading this video. This could be due to a temporary network issue or the video might be unavailable.

WebSockets just landed in TanStack AI. 🚀 One socket for the whole conversation. Full-duplex. And if it drops mid-stream, the client reopens at the last offset and replays only the gap, no second model call. Works on Node (`ws`) and Cloudflare (`toWebSocketResponse`)

TANSTACK

65,644 subscribers

53,716 views • 11 days ago •via X (Twitter)

Science & Technology

Anya Rossi• Live Now

Private livecam show

0 Comments

No comments available

Comments from the original post will appear here

Related Videos

Listen, speak, handle interruptions, and call tools in one live conversation. 🎙️ Introducing NVIDIA-NemotronLabs-VoiceChat-11B, NVIDIA’s end-to-end full-duplex model for real-time voice agents. 🤖 🏆 It ranks #2 among open full-duplex models on both VoiceBench and Full-Duplex-Bench 1.0. ⚡ Natural turn-taking responds in ~448 ms, while barge-in lets users interrupt the model with ~480 ms latency. 🛠️ The first open full-duplex model with live tool calling. One unified system handles speech understanding, generation, transcription, and tool execution while keeping the conversation flowing. 🧠 Built with a Fast Conformer, Nemotron Nano v2, and a TTS decoder, trained on ~550K hours of real and synthetic speech. 📜 Research use only. OpenMDW 1.1.

Listen, speak, handle interruptions, and call tools in one live conversation. 🎙️ Introducing NVIDIA-NemotronLabs-VoiceChat-11B, NVIDIA’s end-to-end full-duplex model for real-time voice agents. 🤖 🏆 It ranks #2 among open full-duplex models on both VoiceBench and Full-Duplex-Bench 1.0. ⚡ Natural turn-taking responds in ~448 ms, while barge-in lets users interrupt the model with ~480 ms latency. 🛠️ The first open full-duplex model with live tool calling. One unified system handles speech understanding, generation, transcription, and tool execution while keeping the conversation flowing. 🧠 Built with a Fast Conformer, Nemotron Nano v2, and a TTS decoder, trained on ~550K hours of real and synthetic speech. 📜 Research use only. OpenMDW 1.1.

ModelScope

22,194 views • 27 days ago

We just released a new version of TanStack AI! 🚀 In today's release we have added client/server context for tools and middleware. Plug it into your useChat/chat and have it available, fully typesafe out of the box. 🔥

We just released a new version of TanStack AI! 🚀 In today's release we have added client/server context for tools and middleware. Plug it into your useChat/chat and have it available, fully typesafe out of the box. 🔥

TANSTACK

33,646 views • 3 months ago

NVIDIA just removed the biggest friction point in Voice AI 🤯 They've open-sourced PersonaPlex-7B, a full-duplex conversational model that can listen and speak at the same time. Instead of waiting for you to finish talking, it uses a dual-stream architecture to process and respond in real-time. 100% Open-Source and Free.

NVIDIA just removed the biggest friction point in Voice AI 🤯 They've open-sourced PersonaPlex-7B, a full-duplex conversational model that can listen and speak at the same time. Instead of waiting for you to finish talking, it uses a dual-stream architecture to process and respond in real-time. 100% Open-Source and Free.

Simplifying AI

278,970 views • 7 months ago

TanStack AI 🤝Angular Today we have released an official Angular adapter for TanStack AI! 🚀 Try it out in your Angular apps today and find the docs here:

TanStack AI 🤝Angular Today we have released an official Angular adapter for TanStack AI! 🚀 Try it out in your Angular apps today and find the docs here:

TANSTACK

11,461 views • 2 months ago

Triple Crown just landed in the lobby - and it only asks one thing of you. 👑 Flip the coin. Chase three crowns - Just you, a coin, and that little breath you hold mid-flip. Three crowns line up? That's the whole feeling right there ✌️

Triple Crown just landed in the lobby - and it only asks one thing of you. 👑 Flip the coin. Chase three crowns - Just you, a coin, and that little breath you hold mid-flip. Three crowns line up? That's the whole feeling right there ✌️

CrownCoins

32,230 views • 11 days ago

What does it take for CTOs and CIOs to build an AI stack where data reaches compute reliably, securely, and at scale? That’s the strategic question behind my conversation with Kunal Anand, CPO at #F5. As AI moves into production, the real differentiator won’t just be the model — it will be the infrastructure that delivers data with reliability, security, and visibility. The last mile of AI is becoming a first-class product decision, not just an infrastructure one. Watch the full conversation: Register for AI Virtual Summit (free) #AI #CTO #CIO #EnterpriseAI #Infrastructure #Cybersecurity #F5Partner

What does it take for CTOs and CIOs to build an AI stack where data reaches compute reliably, securely, and at scale? That’s the strategic question behind my conversation with Kunal Anand, CPO at #F5. As AI moves into production, the real differentiator won’t just be the model — it will be the infrastructure that delivers data with reliability, security, and visibility. The last mile of AI is becoming a first-class product decision, not just an infrastructure one. Watch the full conversation: Register for AI Virtual Summit (free) #AI #CTO #CIO #EnterpriseAI #Infrastructure #Cybersecurity #F5Partner

Helen Yu

12,148 views • 2 months ago

This is a real hack for ordering from a drive thru that uses AI to take your order Instead of ordering the order naturally with pauses in your speech for the different items, just order the entire order as ‘one run on sentence’ This works because AI drive-thru systems get tripped up by the normal back-and-forth pauses of human conversation So if you do your whole order as one run on sentence, the artificial intelligence understands you better ‘By dumping everything into one long, continuous stream, including all customizations, sizes, and add-ons, you give the AI the full context at once’

This is a real hack for ordering from a drive thru that uses AI to take your order Instead of ordering the order naturally with pauses in your speech for the different items, just order the entire order as ‘one run on sentence’ This works because AI drive-thru systems get tripped up by the normal back-and-forth pauses of human conversation So if you do your whole order as one run on sentence, the artificial intelligence understands you better ‘By dumping everything into one long, continuous stream, including all customizations, sizes, and add-ons, you give the AI the full context at once’

Wall Street Apes

191,905 views • 1 month ago

running a 405-billion-parameter model used to mean a $30,000 GPU rack. one developer made it run on a single 8GB card the kind that's been sitting in gaming laptops for years. it's called AirLLM, and the trick is almost stupidly simple: instead of loading the whole model, it keeps one layer on the GPU at a time and streams the rest through. / Llama 3.1 405B on 8GB / DeepSeek-V3, all 671B of it, on about 12GB / no quantization, no distillation, no pruning the full-precision model / works with Llama, Qwen, DeepSeek, Mistral, Gemma, and more it's Apache 2.0 and already past 22,600 stars. → bookmark it.

running a 405-billion-parameter model used to mean a $30,000 GPU rack. one developer made it run on a single 8GB card the kind that's been sitting in gaming laptops for years. it's called AirLLM, and the trick is almost stupidly simple: instead of loading the whole model, it keeps one layer on the GPU at a time and streams the rest through. / Llama 3.1 405B on 8GB / DeepSeek-V3, all 671B of it, on about 12GB / no quantization, no distillation, no pruning the full-precision model / works with Llama, Qwen, DeepSeek, Mistral, Gemma, and more it's Apache 2.0 and already past 22,600 stars. → bookmark it.

dunik

151,882 views • 1 month ago

Generate 10-second videos with native audio in under 10 seconds. P-video by Pruna AI is live on Cloudflare AI Gateway and it's free for the next 24 hours.

Generate 10-second videos with native audio in under 10 seconds. P-video by Pruna AI is live on Cloudflare AI Gateway and it's free for the next 24 hours.

Cloudflare

17,533 views • 6 months ago

What happens when the Internet is no longer mostly used by humans, but by bots, AI agents, and machines? Cloudflare co-founder and CEO Matthew Prince joins Matt Turck on The MAD Podcast to break down the shift to a machine-to-machine web. Watch the full conversation:

What happens when the Internet is no longer mostly used by humans, but by bots, AI agents, and machines? Cloudflare co-founder and CEO Matthew Prince joins Matt Turck on The MAD Podcast to break down the shift to a machine-to-machine web. Watch the full conversation:

Cloudflare

26,515 views • 2 months ago

Burundi special commandos. Over 25 "commandos" and only one landed in the stadium. The rest landed in forests, on top of houses and electric polls! Thanks to the power shortage in the country, no one was electrocuted.

Burundi special commandos. Over 25 "commandos" and only one landed in the stadium. The rest landed in forests, on top of houses and electric polls! Thanks to the power shortage in the country, no one was electrocuted.

Dr. Dash

249,417 views • 10 months ago

SNEAKO stopped at a coffee shop on stream for a drink and asked a worker if he remembered the guy who threatened to spit in it, and the employee said he saw it happen and that the person still works there. 😳

SNEAKO stopped at a coffee shop on stream for a drink and asked a worker if he remembered the guy who threatened to spit in it, and the employee said he saw it happen and that the person still works there. 😳

SNEAKO UPDATES

113,177 views • 6 months ago

I can't believe we're still debating the carbon price. It works and the rebates make the majority of Canadians whole or better off. Listen to the full conversation at the link below:

I can't believe we're still debating the carbon price. It works and the rebates make the majority of Canadians whole or better off. Listen to the full conversation at the link below:

Nate Erskine-Smith

340,543 views • 2 years ago

I plugged ChatGPT Luna 5.6 and Claude Opus 5 into Blender through MCP, and one of them is one-shotting entire 3D scenes now. Top is Luna 5.6 in fast mode modeling ship cannons with a firing animation. Watching the scene grow mesh by mesh in real time is a separate kind of dopamine. Bottom is Opus 5 building a full carriage scene with terrain, rocks and lighting. This one is simply built for 3D -- the whole thing from a single prompt with zero retries and zero corrections. The entire bridge is one open-source repo, no Python and no modeling skills required: > The repo is with 16.9k stars under MIT license. > It works with any model -- Claude, GPT, Cursor, any MCP client. > The model sees your scene through a live socket and edits it in real time. A year ago this workflow didn't exist, and now the gap between an idea and a rendered 3D scene is one prompt. Save this one -- you will come back for the repo ↓

I plugged ChatGPT Luna 5.6 and Claude Opus 5 into Blender through MCP, and one of them is one-shotting entire 3D scenes now. Top is Luna 5.6 in fast mode modeling ship cannons with a firing animation. Watching the scene grow mesh by mesh in real time is a separate kind of dopamine. Bottom is Opus 5 building a full carriage scene with terrain, rocks and lighting. This one is simply built for 3D -- the whole thing from a single prompt with zero retries and zero corrections. The entire bridge is one open-source repo, no Python and no modeling skills required: > The repo is with 16.9k stars under MIT license. > It works with any model -- Claude, GPT, Cursor, any MCP client. > The model sees your scene through a live socket and edits it in real time. A year ago this workflow didn't exist, and now the gap between an idea and a rendered 3D scene is one prompt. Save this one -- you will come back for the repo ↓

slash1s

150,102 views • 1 month ago

On July 22, investors at the top of their game will share their BEST stock ideas with you. One after another, and the whole day costs just $99. These are the people who actually manage money, and you get the replays if you can't make it live. Sign up:

On July 22, investors at the top of their game will share their BEST stock ideas with you. One after another, and the whole day costs just $99. These are the people who actually manage money, and you get the replays if you can't make it live. Sign up:

George Noble

88,601 views • 2 months ago

Will AI enforce laws at the cost of personal freedom? On The RichRollPodcast talking about AI-powered surveillance and the ethical dilemmas it poses for privacy and control. Watch the full conversation here: #AI #Surveillance #AbortionLaws richroll

Will AI enforce laws at the cost of personal freedom? On The RichRollPodcast talking about AI-powered surveillance and the ethical dilemmas it poses for privacy and control. Watch the full conversation here: #AI #Surveillance #AbortionLaws richroll

Yuval Noah Harari

26,339 views • 1 year ago

This Chinese developer linked two $2,999 NVIDIA DGX Sparks into one box and runs the full Qwen3-235B at home, after dropping his $1,999-a-month cloud bill to zero. He wired 2 small boxes into a single computer, split a giant 235-billion-parameter model in half between them, and serves it across his own network at about 10 tokens a second, with no internet, no cloud, right there on the desk. No data center, no thousand-dollar graphics cards, no monthly cloud bill. Just him, 2 gold boxes the size of a sandwich, one cable between them, and 1 power strip. And here is the whole payoff. He used to pay the cloud $1,999 a month for the same model, and the meter ticked on every request. Now he paid $5,998 once for 2 boxes, they covered their cost in 3 months, and after that he sends as many requests as he wants for free, only electricity. The two Sparks talk over one fast cable, each holds 128GB of memory, and together they carry the whole model, about 73GB loaded per box, with the chip inside pinned near the limit at 96%. Both boxes work as one and keep trading data over the cable, with no cloud in the loop and no single word leaking out. The ready model sits on one local address, and any app on his network calls it as easily as ChatGPT. And here is how he described, in plain words, what this pair of boxes does: "this is a pair of boxes that holds the huge Qwen3-235B model and serves it to one network. the model is split in half, and each box owns its half. parts: // Box 1 (holds the first half of the model and starts the answer fast, the first word appears in under a second) // Box 2 (holds the second half and writes out the rest, about 10 tokens a second) // Cable (connects the 2 boxes and moves data between them on every step, with no lag) // Address (one local address where any app sends its request, like to a cloud model) // Test (a script that runs big prompts through and measures speed and delays) // Monitor (checks temperature, power draw, and load on both boxes every 2 seconds). the model never goes to the cloud. he only steps in when a box runs hotter than 80 degrees or the cable between them starts dropping data." So the system knows exactly what it is, what it is for, and where its limits are. It knows it has to hold the whole huge model across 2 boxes on its own. It knows it has to answer every request locally, with no meter, no limits, and no internet. It knows the human is only needed when a box overheats or the link between them stalls. → The setup runs around the clock on 2 boxes, each pulling under 60 watts → However many requests he sends, the monthly bill is $0, only electricity → The first box starts the answer in under a second → The second writes text at about 10 tokens a second → One request at a time: 838 tokens in 85 seconds, first word in 0.8s → Two requests at once: 697 tokens in 108 seconds, first word in 0.7s → Both boxes sit at 96% load and warm up to 76-78 degrees And only when a chip in a box runs hotter than 80 degrees or the cable between the 2 Sparks drops data does the system call the owner. And when he himself is out on a run or in a coffee shop, he still reaches his own model at home from his phone: sends a big prompt to the local Qwen3-235B, gets the full answer back in under a minute and a half, with no token meter ticking and no limit to hit. Here is what the test shows on his screen during one of the night runs: "one request at a time: 838 tokens in 84.9 seconds, first word in 0.8s, then 0.1s per token." "two requests at once: 697 tokens in 107.6 seconds, first word in 0.7s, then 0.15s per token." "Box 1: chip at 96% load, 76 degrees, 56 watts, 73GB used in memory." "Box 2: chip at 96% load, 78 degrees, 56 watts, the Qwen3-235B model fully loaded." And while everyone around is paying for AI by the month and bumping into limits, his top-tier model just sits on the desk and works as much as he wants: his own little power plant instead of a forever meter. He has no server rack of his own and no cloud account behind it. Just 2 DGX Spark boxes on a desk, one model split in half between them, one local address, and a folder of prompts next to it. Out of everything I have seen this year, this is the cleanest way to stop paying for AI: $5,998 of hardware on the desk once, $0 a month to the cloud, unlimited forever, and between them 2 gold boxes, 1 cable, and the full Qwen3-235B answering at home with no internet.

This Chinese developer linked two $2,999 NVIDIA DGX Sparks into one box and runs the full Qwen3-235B at home, after dropping his $1,999-a-month cloud bill to zero. He wired 2 small boxes into a single computer, split a giant 235-billion-parameter model in half between them, and serves it across his own network at about 10 tokens a second, with no internet, no cloud, right there on the desk. No data center, no thousand-dollar graphics cards, no monthly cloud bill. Just him, 2 gold boxes the size of a sandwich, one cable between them, and 1 power strip. And here is the whole payoff. He used to pay the cloud $1,999 a month for the same model, and the meter ticked on every request. Now he paid $5,998 once for 2 boxes, they covered their cost in 3 months, and after that he sends as many requests as he wants for free, only electricity. The two Sparks talk over one fast cable, each holds 128GB of memory, and together they carry the whole model, about 73GB loaded per box, with the chip inside pinned near the limit at 96%. Both boxes work as one and keep trading data over the cable, with no cloud in the loop and no single word leaking out. The ready model sits on one local address, and any app on his network calls it as easily as ChatGPT. And here is how he described, in plain words, what this pair of boxes does: "this is a pair of boxes that holds the huge Qwen3-235B model and serves it to one network. the model is split in half, and each box owns its half. parts: // Box 1 (holds the first half of the model and starts the answer fast, the first word appears in under a second) // Box 2 (holds the second half and writes out the rest, about 10 tokens a second) // Cable (connects the 2 boxes and moves data between them on every step, with no lag) // Address (one local address where any app sends its request, like to a cloud model) // Test (a script that runs big prompts through and measures speed and delays) // Monitor (checks temperature, power draw, and load on both boxes every 2 seconds). the model never goes to the cloud. he only steps in when a box runs hotter than 80 degrees or the cable between them starts dropping data." So the system knows exactly what it is, what it is for, and where its limits are. It knows it has to hold the whole huge model across 2 boxes on its own. It knows it has to answer every request locally, with no meter, no limits, and no internet. It knows the human is only needed when a box overheats or the link between them stalls. → The setup runs around the clock on 2 boxes, each pulling under 60 watts → However many requests he sends, the monthly bill is $0, only electricity → The first box starts the answer in under a second → The second writes text at about 10 tokens a second → One request at a time: 838 tokens in 85 seconds, first word in 0.8s → Two requests at once: 697 tokens in 108 seconds, first word in 0.7s → Both boxes sit at 96% load and warm up to 76-78 degrees And only when a chip in a box runs hotter than 80 degrees or the cable between the 2 Sparks drops data does the system call the owner. And when he himself is out on a run or in a coffee shop, he still reaches his own model at home from his phone: sends a big prompt to the local Qwen3-235B, gets the full answer back in under a minute and a half, with no token meter ticking and no limit to hit. Here is what the test shows on his screen during one of the night runs: "one request at a time: 838 tokens in 84.9 seconds, first word in 0.8s, then 0.1s per token." "two requests at once: 697 tokens in 107.6 seconds, first word in 0.7s, then 0.15s per token." "Box 1: chip at 96% load, 76 degrees, 56 watts, 73GB used in memory." "Box 2: chip at 96% load, 78 degrees, 56 watts, the Qwen3-235B model fully loaded." And while everyone around is paying for AI by the month and bumping into limits, his top-tier model just sits on the desk and works as much as he wants: his own little power plant instead of a forever meter. He has no server rack of his own and no cloud account behind it. Just 2 DGX Spark boxes on a desk, one model split in half between them, one local address, and a folder of prompts next to it. Out of everything I have seen this year, this is the cleanest way to stop paying for AI: $5,998 of hardware on the desk once, $0 a month to the cloud, unlimited forever, and between them 2 gold boxes, 1 cable, and the full Qwen3-235B answering at home with no internet.

Blaze

93,871 views • 3 months ago

Vaccinate your kids." Four words dropped mid-conversation just detonated the entire debate — and what came right after is NOT what anyone expected. It starts calm, almost too calm, and then in one single sentence the whole energy flips. By the halfway mark the tension is so thick you can feel it through the screen, and the final response is the kind of line that gets clipped, quoted, and argued about in the comments for days. Watch all the way through — the last ten seconds hit different. Watch the full episode on my YT channel! Link is in the comments 📷

Vaccinate your kids." Four words dropped mid-conversation just detonated the entire debate — and what came right after is NOT what anyone expected. It starts calm, almost too calm, and then in one single sentence the whole energy flips. By the halfway mark the tension is so thick you can feel it through the screen, and the final response is the kind of line that gets clipped, quoted, and argued about in the comments for days. Watch all the way through — the last ten seconds hit different. Watch the full episode on my YT channel! Link is in the comments 📷

Jillian Michaels

36,245 views • 13 days ago