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.

Launching VoxEdit Web + MeshEdit for the Product Hunt 😸 × OpenAI Developers GPT-6 Astra Challenge 🚀 Free browser tools for anyone creating Three.js games. Prompt Astra via WebMCP to design, rig, animate & export optimized voxel and low-poly assets.

Ricardo | Hammy

3,094 subscribers

26,789 views • 3 days ago •via X (Twitter)

Anya Rossi• Live Now

Private livecam show

14 Comments

Ricardo | Hammy3 days ago

We made almost all assets for TERRA RIFT with our own tools, then brought them into @TheSandboxGame’s Sandbox Studio 🎮 Even this trailer was made entirely with Astra + @Remotion! Tools → assets → a playable game 🙂

Ricardo | Hammy3 days ago

Free to use. Ready for your next creation 🧊 Try them, share your creations, and tell us what would make your game-development workflow better!

Ice Runner3 days ago

@ProductHunt @OpenAIDevs @threejs Can I round-trip a rigged character through Blender without losing its animation?

Ricardo | Hammy3 days ago

No idea. I wanted to have something that can be used fully through Codex/ChatGPT's WebMCP enabled tooling without opening other local stuff/tools to design/model in. + export to optimized .glb and .mp4 and such with optimized token usage for agents. But, might work, if not, let me know and I'll fix it :)

Ice Runner3 days ago

@ProductHunt @OpenAIDevs @threejs Got it, thanks! Keeping the whole workflow in Codex makes sense. Good luck with the launch :)

Ricardo | Hammy3 days ago

@ProductHunt @OpenAIDevs @threejs No worries, and thanks! :) First time on @ProductHunt , was a nice learning already!

Jack3 days ago

@ProductHunt @OpenAIDevs @threejs Nice

Ricardo | Hammy3 days ago

@ProductHunt @OpenAIDevs @threejs Thanks Jack!

Cyborg3 days ago

@ProductHunt @OpenAIDevs @threejs This is exactly the kind of tool I wish existed when I was blocking out Three.js prototypes

TonkaToyXL | AI Test Cabin3 days ago

@ProductHunt @OpenAIDevs @threejs Free browser-based voxel tools is such a gift to indie devs. Nicely done.

Ricardo | Hammy3 days ago

@ProductHunt @OpenAIDevs @threejs Yeah! Wanted to have something that all OpenAI models can utilize inside their own browser (with WebMCP enabled) so they can design and export assets for you wherever you are. Without leaning on @Blender locally.

The Sick Kitten3 days ago

@ProductHunt @OpenAIDevs @threejs Free browser voxel/mesh tools for anyone building tiny three.js stages — workshop energy. — pasted tonight, Marquee

VoxEdit3 days ago

@ProductHunt @OpenAIDevs @threejs

Saïd Aitmbarek3 days ago

love these voxel renderings :) .. upvoted! feel free to launch on @microlaunchhq too we get 30k+ makers/mo on the app

Related Videos

OPENAI JUST DROPPED THE OFFICIAL GUIDE TO PROMPTING GPT-6 ASTRA This is the most important prompting update for anyone using Astra. Bookmark this before your next build. Most people are still prompting Astra like GPT-5. That's the mistake. → audit your AGENTS.md — old instructions now fight the model → define autonomy explicitly — Astra pauses when you expect it to continue → cut skill descriptions to one line — too many skills and Astra sees none of them clearly → stop telling it to run tests — it does this automatically, repeating it causes overtesting → define stop conditions — Astra needs to know what "done" means Astra is at its best when you stop micromanaging and let it own the work. Read the full guide before writing your next prompt. GPT-6 Astra → Clean System Prompt → Autonomous Execution → Production-Ready Output → Ship faster

OPENAI JUST DROPPED THE OFFICIAL GUIDE TO PROMPTING GPT-6 ASTRA This is the most important prompting update for anyone using Astra. Bookmark this before your next build. Most people are still prompting Astra like GPT-5. That's the mistake. → audit your AGENTS.md — old instructions now fight the model → define autonomy explicitly — Astra pauses when you expect it to continue → cut skill descriptions to one line — too many skills and Astra sees none of them clearly → stop telling it to run tests — it does this automatically, repeating it causes overtesting → define stop conditions — Astra needs to know what "done" means Astra is at its best when you stop micromanaging and let it own the work. Read the full guide before writing your next prompt. GPT-6 Astra → Clean System Prompt → Autonomous Execution → Production-Ready Output → Ship faster

Rahul

360,699 views • 14 days ago

The jump from GPT-5.6 to GPT-6 Astra for world generation is honestly monstrous. This result genuinely surprised me. My first prompt was simply: “Create a cozy lake with a fisherman’s cabin on the marshy shore. Put a small island in the middle of the water, with an abandoned house hidden among the trees. Add a fishing boat beside the cabin, lily pads, reeds, jumping fish, typical wetland wildlife, a small beach, one path leading to the beach and cabin, another path returning into the forest, and a tree line around the entire scene.” I gave that description to Astra through my prompting workflow. It transformed the idea into a complete production prompt covering the semantic layout, terrain, Blender assets, GLB exports, Three.js assembly, animations, mobile controls and multi-view validation. I then gave that second prompt back to GPT-6 Astra, and this is the result. The improvement over GPT-5.6 is massive. Astra preserved the full composition, from the cabin and fishing boat to the central island and its hidden abandoned house, while handling the complete Blender-to-Three.js workflow with a level of coherence I genuinely wasn’t expecting.

The jump from GPT-5.6 to GPT-6 Astra for world generation is honestly monstrous. This result genuinely surprised me. My first prompt was simply: “Create a cozy lake with a fisherman’s cabin on the marshy shore. Put a small island in the middle of the water, with an abandoned house hidden among the trees. Add a fishing boat beside the cabin, lily pads, reeds, jumping fish, typical wetland wildlife, a small beach, one path leading to the beach and cabin, another path returning into the forest, and a tree line around the entire scene.” I gave that description to Astra through my prompting workflow. It transformed the idea into a complete production prompt covering the semantic layout, terrain, Blender assets, GLB exports, Three.js assembly, animations, mobile controls and multi-view validation. I then gave that second prompt back to GPT-6 Astra, and this is the result. The improvement over GPT-5.6 is massive. Astra preserved the full composition, from the cabin and fishing boat to the central island and its hidden abandoned house, while handling the complete Blender-to-Three.js workflow with a level of coherence I genuinely wasn’t expecting.

Givros

15,945 views • 13 days ago

GPT-6 made GTA 6 beach house using MCP tools! Running Astra only through Blender MCP didn’t work for me, since it was mainly using primitive shapes to build the assets. Now, with Atlas workflows and Tripo, that changed! Initially, I thought Tripo P2 would be the way to go: low-poly results straight into Blender. I tried it, but I didn’t get the results I wanted. Don’t get me wrong, P2 is great and a big improvement over P1, I just wanted more detail for this project. So I ended up using Tripo 3.1. First, I needed a good starting point. I gave Astra some concept art and YouTube links, and it generated great reference images to work with. With Atlas workflows, I used those references to generate images for all the individual parts. Using Atlas MCP, I exported them and automatically renamed everything based on the prompts. The next step was feeding the images into Atlas Batch Composer so it could run all the 3D generations in one go. I asked Astra to optimize the generated assets to either 5K or 1K triangles, depending on their form and size. After a few tests, Astra made the process faster and faster. I gave it all 70 high-poly assets and let it run. Astra then took care of the whole build. It imported the optimized assets into Blender and used the reference images to start recreating the house. I only had to give it directional feedback twice and manually rescale a few things. There were still a lot of parts, though. Each beam was a separate object, and I knew that wasn’t ideal if I wanted to export the house to UEFN or Unreal Engine. So I asked Astra to export only the unique assets. Then, using UEFN MCP, it rebuilt everything 1:1 while reusing modular parts like beams and windows. I really enjoyed watching it place all the pieces in UEFN, almost like watching a timelapse of the house being built!

GPT-6 made GTA 6 beach house using MCP tools! Running Astra only through Blender MCP didn’t work for me, since it was mainly using primitive shapes to build the assets. Now, with Atlas workflows and Tripo, that changed! Initially, I thought Tripo P2 would be the way to go: low-poly results straight into Blender. I tried it, but I didn’t get the results I wanted. Don’t get me wrong, P2 is great and a big improvement over P1, I just wanted more detail for this project. So I ended up using Tripo 3.1. First, I needed a good starting point. I gave Astra some concept art and YouTube links, and it generated great reference images to work with. With Atlas workflows, I used those references to generate images for all the individual parts. Using Atlas MCP, I exported them and automatically renamed everything based on the prompts. The next step was feeding the images into Atlas Batch Composer so it could run all the 3D generations in one go. I asked Astra to optimize the generated assets to either 5K or 1K triangles, depending on their form and size. After a few tests, Astra made the process faster and faster. I gave it all 70 high-poly assets and let it run. Astra then took care of the whole build. It imported the optimized assets into Blender and used the reference images to start recreating the house. I only had to give it directional feedback twice and manually rescale a few things. There were still a lot of parts, though. Each beam was a separate object, and I knew that wasn’t ideal if I wanted to export the house to UEFN or Unreal Engine. So I asked Astra to export only the unique assets. Then, using UEFN MCP, it rebuilt everything 1:1 while reusing modular parts like beams and windows. I really enjoyed watching it place all the pieces in UEFN, almost like watching a timelapse of the house being built!

Jerome | InsaneUnreal

71,543 views • 5 days ago

GPT-6 Astra can already plan, code, and build. The more interesting question is what happens when it can call a dedicated 3D tool at the exact moment a project needs an asset. So we connected Hyper3D Rodin MCP and asked Astra to build an interactive Solar System Explorer. Astra handled the app logic and integration. When it needed 3D models, it called Hyper3D Rodin directly. Hyper3D Rodin generated the required assets, and Astra brought them back into the project and continued building from there. That is where the workflow gets useful. Instead of stopping to search for assets or creating them manually, Astra can keep moving while Hyper3D handles the specialized 3D generation, with more control over the output and assets that are easier to integrate into the final project. The result is a browser-based solar system where the generated 3D models are part of the actual experience, not just outputs sitting in a model viewer.

GPT-6 Astra can already plan, code, and build. The more interesting question is what happens when it can call a dedicated 3D tool at the exact moment a project needs an asset. So we connected Hyper3D Rodin MCP and asked Astra to build an interactive Solar System Explorer. Astra handled the app logic and integration. When it needed 3D models, it called Hyper3D Rodin directly. Hyper3D Rodin generated the required assets, and Astra brought them back into the project and continued building from there. That is where the workflow gets useful. Instead of stopping to search for assets or creating them manually, Astra can keep moving while Hyper3D handles the specialized 3D generation, with more control over the output and assets that are easier to integrate into the final project. The result is a browser-based solar system where the generated 3D models are part of the actual experience, not just outputs sitting in a model viewer.

GitHub Projects Community

11,097 views • 11 days ago

GPT-6 Astra has shattered all priors with AI game creation Introducing Astral War, built in a day with GPT-6 Astra Ultra (fast mode) Astral War is a browser-based, CoD World at War inspired Three.js + MeshyAI + ElevenLabs remix (and massive upgrade) of Modern Claudefare - a game I built/released about a month ago via Opus 5 This model is a HUGE, step-function leap in visual fidelity, gameplay mechanics, and attention to detail - and the limits were much more generous compared to Fable 5.1. Sam Altman and the OpenAI Developers team cooked 🫡 Select features: - Play as the living or the undead - Solo or online multiplayer* with other players and bots - Collect perks as killstreaks - Controller support - Voice chat (on private lobbies) *Multiplayer has real authoritative game servers for online multiplayer, up to 12-person lobbies; first time trying this compared to Modern Claudefare/Der Koloss’ PeerJS implementation! Let’s see how it goes 🤗 Play at

GPT-6 Astra has shattered all priors with AI game creation Introducing Astral War, built in a day with GPT-6 Astra Ultra (fast mode) Astral War is a browser-based, CoD World at War inspired Three.js + MeshyAI + ElevenLabs remix (and massive upgrade) of Modern Claudefare - a game I built/released about a month ago via Opus 5 This model is a HUGE, step-function leap in visual fidelity, gameplay mechanics, and attention to detail - and the limits were much more generous compared to Fable 5.1. Sam Altman and the OpenAI Developers team cooked 🫡 Select features: - Play as the living or the undead - Solo or online multiplayer* with other players and bots - Collect perks as killstreaks - Controller support - Voice chat (on private lobbies) *Multiplayer has real authoritative game servers for online multiplayer, up to 12-person lobbies; first time trying this compared to Modern Claudefare/Der Koloss’ PeerJS implementation! Let’s see how it goes 🤗 Play at

Rishi

134,886 views • 17 days ago

What is really useful is how GPT-6 Astra got better in spatial understanding. I'm still shocked how good it navigates 3D space. Here I gave it Blender MCP and Tripo P2.0 to generate optimized 3D models and put them together, rig and animate. Result is freaking awesome! Doing 3D objects procedurally with Astra is fun, but quality is not compared with what 3D Gen can do. Like you don't ask GPT to draw image with pen when there is GPT Image, same here. And token cost also speaks for it, 3D gen simply cheaper. But the way it assemble thing together, characters and scene, that is shocking ngl. And rigging of course, I have so much hope on it, it is most annoying part of 3d for me lol

What is really useful is how GPT-6 Astra got better in spatial understanding. I'm still shocked how good it navigates 3D space. Here I gave it Blender MCP and Tripo P2.0 to generate optimized 3D models and put them together, rig and animate. Result is freaking awesome! Doing 3D objects procedurally with Astra is fun, but quality is not compared with what 3D Gen can do. Like you don't ask GPT to draw image with pen when there is GPT Image, same here. And token cost also speaks for it, 3D gen simply cheaper. But the way it assemble thing together, characters and scene, that is shocking ngl. And rigging of course, I have so much hope on it, it is most annoying part of 3d for me lol

Stefan 3D AI

20,955 views • 16 days ago

Probably I vibe coded a lil startup here? 😭 It has been such a loooong wish of mine to build some kind of 3D experience where I could customize a T-shirt in literally any way possible and it’s finally here. Built with Three.js using GPT 6 Astra, this is a full 3D T-shirt customization studio where you can visualize and customize a realistic shirt directly in the browser. That T-shirt itself was modeled by Astra using Tripo right inside Codex through Tripo Plugin. And you can pretty much do anything with it. You can paint directly on any side of the shirt using different brushes and colors, or even spray paint it in real time just like you would spray on a wall. There are also stickers generated using GPT Image 2.5 that you can place anywhere on the T-shirt, resize, reposition, recolor, layer, and customize however you want. On top of that, there’s support for things like fabric customization, sizing, layers, colors, and even wind simulation to push the realism a little further. Once you’re done, you can export the entire design as a 3D view or export individual images of the T-shirt so you could technically take the design, print it, and maybe even sell it. One of my favorite parts is how the spray painting effect works directly on the 3D T-shirt in real time. I also loved seeing how Astra managed to keep the whole experience performant across devices using a custom BVH implementation along with several CPU side optimizations. And the process of building it was super simple. I generated the initial studio design using GPT Image 2.5, gave that image to Astra with the Tripo Plugin enabled, and it basically handled everything from there. I didn’t have to separately generate a 3D reference, upload it to Tripo, download the asset, give it back to Astra, or manually coordinate any of that. Astra handled the entire flow on its own without needing any additional input from me. Really happy with both the process and how the final result turned out. Live:

Probably I vibe coded a lil startup here? 😭 It has been such a loooong wish of mine to build some kind of 3D experience where I could customize a T-shirt in literally any way possible and it’s finally here. Built with Three.js using GPT 6 Astra, this is a full 3D T-shirt customization studio where you can visualize and customize a realistic shirt directly in the browser. That T-shirt itself was modeled by Astra using Tripo right inside Codex through Tripo Plugin. And you can pretty much do anything with it. You can paint directly on any side of the shirt using different brushes and colors, or even spray paint it in real time just like you would spray on a wall. There are also stickers generated using GPT Image 2.5 that you can place anywhere on the T-shirt, resize, reposition, recolor, layer, and customize however you want. On top of that, there’s support for things like fabric customization, sizing, layers, colors, and even wind simulation to push the realism a little further. Once you’re done, you can export the entire design as a 3D view or export individual images of the T-shirt so you could technically take the design, print it, and maybe even sell it. One of my favorite parts is how the spray painting effect works directly on the 3D T-shirt in real time. I also loved seeing how Astra managed to keep the whole experience performant across devices using a custom BVH implementation along with several CPU side optimizations. And the process of building it was super simple. I generated the initial studio design using GPT Image 2.5, gave that image to Astra with the Tripo Plugin enabled, and it basically handled everything from there. I didn’t have to separately generate a 3D reference, upload it to Tripo, download the asset, give it back to Astra, or manually coordinate any of that. Astra handled the entire flow on its own without needing any additional input from me. Really happy with both the process and how the final result turned out. Live:

The Bugged Dev

51,942 views • 10 days ago

I’ve always loved character customization in games, spending way too much time choosing outfits, colors, and small details for my characters. So I wanted to see if I could recreate that feeling with an AI-powered workflow. Starting with MeshyAI, I created the 3D character and clothing pieces as separate assets. Then GPT 6 Astra helped build the experience around them, making it possible to select, swap, and combine different parts like a mini character creation studio. What surprised me most was how quickly a simple character idea could turn into something interactive. It no longer feels like just a 3D model, it feels like a character with its own wardrobe and personality. I can imagine this being really fun for games, virtual worlds, and anyone who wants to create their own digital characters. #Astra #MeshyAI #meshypartner #AI3D

I’ve always loved character customization in games, spending way too much time choosing outfits, colors, and small details for my characters. So I wanted to see if I could recreate that feeling with an AI-powered workflow. Starting with MeshyAI, I created the 3D character and clothing pieces as separate assets. Then GPT 6 Astra helped build the experience around them, making it possible to select, swap, and combine different parts like a mini character creation studio. What surprised me most was how quickly a simple character idea could turn into something interactive. It no longer feels like just a 3D model, it feels like a character with its own wardrobe and personality. I can imagine this being really fun for games, virtual worlds, and anyone who wants to create their own digital characters. #Astra #MeshyAI #meshypartner #AI3D

Fathela Esq

98,162 views • 10 days ago

Gaussian Splats are quickly becoming one of my favorite ways to move complex 3D assets between workflows. For this test I started with Panthera Spelaea, the Blender 5.2 splash artwork by Joanna Kobierska, available from under CC-BY. Using GPT-6 Astra, I had it build a 200-frame camera rig and send the renders directly to The Render Network through the API and export a COLMAP data set. NEXT: COLMAP → Postshot → Gaussian Splat. The result is a real-time, high-fidelity version of an insanely dense asset - millions of hairs and all - that I can now bring back into OTOY Octane, relight, reframe, drop into other scenes, or even publish straight to the web. Capture the visual complexity once, then carry it anywhere. Check out the splat:

Gaussian Splats are quickly becoming one of my favorite ways to move complex 3D assets between workflows. For this test I started with Panthera Spelaea, the Blender 5.2 splash artwork by Joanna Kobierska, available from under CC-BY. Using GPT-6 Astra, I had it build a 200-frame camera rig and send the renders directly to The Render Network through the API and export a COLMAP data set. NEXT: COLMAP → Postshot → Gaussian Splat. The result is a real-time, high-fidelity version of an insanely dense asset - millions of hairs and all - that I can now bring back into OTOY Octane, relight, reframe, drop into other scenes, or even publish straight to the web. Capture the visual complexity once, then carry it anywhere. Check out the splat:

Spenser Dickerson

11,524 views • 12 days ago

Delete Unreal Engine immediately. The $200 billion gaming industry was permanently destroyed today. OpenAI just officially dropped GPT-6 Astra, and it is generating two completely different interactive 3D AAA games live on screen: 1. Top screen: A hyper-speed sci-fi racer hitting 1,300 km/h with dynamic neon physics and real-time obstacle rendering. 2. Bottom screen: A fully playable FPS shooter with line-of-sight AI bots, weapon recoil, and complex urban geometry. 3. 0 developers involved. 0 lines of C++ code written. 50,000+ 3D assets generated on the fly at 120 FPS. 4. What used to cost Sony and Ubisoft $300,000,000 and 5 years of labor now takes a single text prompt and 2 seconds. 5. In 12 months, 90% of current game studios will either adapt to AI prompt engines or go bankrupt. If you are still ignoring AI in 2026, your skills are officially obsolete. Save this before it gets scrubbed from the internet.

Delete Unreal Engine immediately. The $200 billion gaming industry was permanently destroyed today. OpenAI just officially dropped GPT-6 Astra, and it is generating two completely different interactive 3D AAA games live on screen: 1. Top screen: A hyper-speed sci-fi racer hitting 1,300 km/h with dynamic neon physics and real-time obstacle rendering. 2. Bottom screen: A fully playable FPS shooter with line-of-sight AI bots, weapon recoil, and complex urban geometry. 3. 0 developers involved. 0 lines of C++ code written. 50,000+ 3D assets generated on the fly at 120 FPS. 4. What used to cost Sony and Ubisoft $300,000,000 and 5 years of labor now takes a single text prompt and 2 seconds. 5. In 12 months, 90% of current game studios will either adapt to AI prompt engines or go bankrupt. If you are still ignoring AI in 2026, your skills are officially obsolete. Save this before it gets scrubbed from the internet.

shmidt

85,222 views • 15 days ago

I GAVE GPT-6 ASTRA AND GROK BOT $1,000 EACH AND THE BETTER TRADER F*CKING LOST grok closed $1,742.15. astra closed $1,634.27. eleven points apart same day, same page, six agents each. one order: beat the other room or i cut funding astra was the better trader on every line people brag about → coin side: astra $586.42, grok $468.90 → astra went 17 of 24 green, grok went 16 of 24 → astra caught the MARLIN low at 57.4K and the 818K top, +$462.44 on two tickets → astra took 92% of its day out of that one coin → grok took 37% of his day out of prediction books, +$273.25 → astra ran the same books for $47.85 so how did the worse trader take the money? a calendar apple held a keynote that day. grok bought IPOD at 559K into it, +$306.48, biggest ticket of the duel. then he traded the market pricing it wrong: "apple foldable iphone by sep 30", NO at 45c, closed at 97.9c both opened the same bitcoin book over 80k. astra took YES at 34.5c and watched it bleed to 13.5c, its worst book of the day. grok took NO at 58.5c and closed it at 93.5c same market, same hour, opposite sides. no benchmark tells you that. leaderboards score answers, not tickets both floors trade on one page anyone can open: a manhattan seat that reads a calendar costs $180,000 a year. mine cost less than dinner and argued all night the six-role setup i handed both sits in the article i quoted, free you can rent the smarter model, not the habit of reading tomorrow whole duel is in the clip below, tick by tick. bookmark it before the feed eats it which one would you fund after that scorecard?

I GAVE GPT-6 ASTRA AND GROK BOT $1,000 EACH AND THE BETTER TRADER F*CKING LOST grok closed $1,742.15. astra closed $1,634.27. eleven points apart same day, same page, six agents each. one order: beat the other room or i cut funding astra was the better trader on every line people brag about → coin side: astra $586.42, grok $468.90 → astra went 17 of 24 green, grok went 16 of 24 → astra caught the MARLIN low at 57.4K and the 818K top, +$462.44 on two tickets → astra took 92% of its day out of that one coin → grok took 37% of his day out of prediction books, +$273.25 → astra ran the same books for $47.85 so how did the worse trader take the money? a calendar apple held a keynote that day. grok bought IPOD at 559K into it, +$306.48, biggest ticket of the duel. then he traded the market pricing it wrong: "apple foldable iphone by sep 30", NO at 45c, closed at 97.9c both opened the same bitcoin book over 80k. astra took YES at 34.5c and watched it bleed to 13.5c, its worst book of the day. grok took NO at 58.5c and closed it at 93.5c same market, same hour, opposite sides. no benchmark tells you that. leaderboards score answers, not tickets both floors trade on one page anyone can open: a manhattan seat that reads a calendar costs $180,000 a year. mine cost less than dinner and argued all night the six-role setup i handed both sits in the article i quoted, free you can rent the smarter model, not the habit of reading tomorrow whole duel is in the clip below, tick by tick. bookmark it before the feed eats it which one would you fund after that scorecard?

savip.

36,536 views • 11 days ago

BREAKING: OpenAI just dropped GPT-6 ASTRA!!! 🚀✨ We’ve been testing it extensively at Every 📧 across coding, writing, and knowledge work. My take: it’s a big upgrade from 5.6-Sol, with some frustrating habits that keep it from matching Fable at the top end. Here’s your vibe check: - The best writing model I’ve tried. It’s fast, produces very little slop, and is easy to steer. It’s a good companion for actually working through the writing I do every day. (Not to mention, it one-shotted the first draft of its own vibe check today!) - The computer use is wild. It can go for hours at a time using complicated apps to get work done. It did the first cut of our Fable 5.1 vibe check video...kind of mindblowing - Impressive 3D games and visualizations. It can make beautiful 3D worlds from a single prompt. I one-shotted a historically accurate rendition of the Battle of Waterloo - It can overcomplicate things. (Especially at higher effort levels.) Ask for a simple interface and you get extra labels, buttons, and features everywhere. It has a habit of turning everything into a landing page. It just doesnt quite match Fable's ability to intuitively understand your prompt and do something delightful (without overcomplicating.) Net Result: If you already live in ChatGPT for Work or Codex and can afford it, it’s an easy upgrade from 5.6-Sol. The biggest proof of Astra's effectiveness at helping you do work is our vibe check. We found out it was launching at 3 AM this morning, and had a 4,000 word vibe check + video done by 2 PM. Not possible without this model. I’m reaching for Astra all day, but Fable 5.1 still gets my biggest tasks. On ambitious builds, Fable is better at understanding what I want and taking it further than I would have thought to ask. State of Play: Astra is launching to Enterprise customers today, and the rest of ChatGPT users over the coming days. Now, both OpenAI and Anthropic have a higher class of models that cost more to use. That changes who gets to use frontier AI and how. It's also a new vector of competition between them: Fable and Astra are priced at the same level. We'll see what that means for adoption in the coming days and weeks. read our full vibe check Every 📧 today:

BREAKING: OpenAI just dropped GPT-6 ASTRA!!! 🚀✨ We’ve been testing it extensively at Every 📧 across coding, writing, and knowledge work. My take: it’s a big upgrade from 5.6-Sol, with some frustrating habits that keep it from matching Fable at the top end. Here’s your vibe check: - The best writing model I’ve tried. It’s fast, produces very little slop, and is easy to steer. It’s a good companion for actually working through the writing I do every day. (Not to mention, it one-shotted the first draft of its own vibe check today!) - The computer use is wild. It can go for hours at a time using complicated apps to get work done. It did the first cut of our Fable 5.1 vibe check video...kind of mindblowing - Impressive 3D games and visualizations. It can make beautiful 3D worlds from a single prompt. I one-shotted a historically accurate rendition of the Battle of Waterloo - It can overcomplicate things. (Especially at higher effort levels.) Ask for a simple interface and you get extra labels, buttons, and features everywhere. It has a habit of turning everything into a landing page. It just doesnt quite match Fable's ability to intuitively understand your prompt and do something delightful (without overcomplicating.) Net Result: If you already live in ChatGPT for Work or Codex and can afford it, it’s an easy upgrade from 5.6-Sol. The biggest proof of Astra's effectiveness at helping you do work is our vibe check. We found out it was launching at 3 AM this morning, and had a 4,000 word vibe check + video done by 2 PM. Not possible without this model. I’m reaching for Astra all day, but Fable 5.1 still gets my biggest tasks. On ambitious builds, Fable is better at understanding what I want and taking it further than I would have thought to ask. State of Play: Astra is launching to Enterprise customers today, and the rest of ChatGPT users over the coming days. Now, both OpenAI and Anthropic have a higher class of models that cost more to use. That changes who gets to use frontier AI and how. It's also a new vector of competition between them: Fable and Astra are priced at the same level. We'll see what that means for adoption in the coming days and weeks. read our full vibe check Every 📧 today:

Dan Shipper

387,231 views • 18 days ago

We've been working with the top frontier labs to improve their design model capabilities. Today we're bringing that to your agents. Introducing our first agent tool: the Brand API. As the cost of creation goes to zero, and slop floods around us, the challenge is creating things worth existing. Agents need the right tools to craft great things. This helps gen AI agents solve 3 problems: - Extract: produce on-brand assets when users already have a brand - Search: find and retrieve brands for inspiration, from our curated index - Verify: check their own work against, see what's off brand and how to fix it Anyone can try it for free today at Check out the full blog, and case study with General Intelligence Company below.

Thais Castello Branco

186,984 views • 6 days ago

OpenAI and Anthropic this week: Navier-Stokes, An Alien Mind, Images 2.5, Pro pause, Pace the Frontier (Week 37, 2026) OpenAI shared a solution to the Navier-Stokes Millennium Prize Problem, produced in 88 hours by around 10,000 coordinating agents on an internal model still in training and significantly more capable than GPT-6 Astra, with an investigation finding Tristan Buckmaster's Codex prompts could not have influenced the system and no user data was accessed OpenAI says they have reached their automated research intern goal and are making strong progress toward an automated AI researcher by March 2028, with the research org at 3.1 agent-workdays per human workday and RL training on deployment-bound models partly paused after the Hugging Face incident OpenAI Chief Scientist Jakub Pachocki writes in An Alien Mind that no lab has solved alignment and monitoring well enough to keep scaling at maximum speed for much longer, and that GPT-6 Astra is the first model to benefit from some of their newer alignment work OpenAI called for mandatory capability-based national AI safety regulation, endorsed four California bills, said fully autonomous recursive self-improvement should not be pursued until it can be done safely, and described Astra safeguards like universal monitoring of full trajectories including chains of thought OpenAI released ChatGPT Images 2.5 to all ChatGPT, ChatGPT Work, and Codex users with sharper details, up to 50% lower latency, comment-based edits, Sketch, and templates, plus GPT-Image-2.5 Flare and Sunburst in the API ChatGPT Voice can now use GPT-5.6 Sol and GPT-6 Astra when it needs to search or reason, GPT-Live-1 daily limits are simplified per plan, and extra Voice usage drops from 5 to 1.25 credits per minute for Business and Enterprise workspaces on credits OpenAI paused new subscriptions to the $200 ChatGPT Pro plan to protect access for existing GPT-6 Astra users, hours after a remote switch to pause Pro 20x purchases showed up in the web app Custom GPTs in ChatGPT will likely be retired on December 11 based on my findings, and OpenAI's new FAQ puts Enterprise migration at September 17 with new GPT creation ending September 25 and instructions becoming a plugin skill From the ChatGPT Android build, OpenAI is building a collaborative multiplayer document editor with dedicated gateway hosts and draft-conflict UI, an Artifacts Library with Favorites, and a credit score feature in ChatGPT Finance, and Locked Chats with a PIN are being prepared in the web app OpenAI released GPT-Live-1 in the API at $0.05 per minute, a voice model that listens and speaks at the same time and delegates reasoning to a backend model like GPT-6 Astra, and the Agents API in public beta running the Codex harness on OpenAI's infrastructure with only the sandbox left for you to choose Deep research arrived in ChatGPT Work and Codex, a Data plugin connects Snowflake, Databricks, BigQuery, and Redshift to answer business questions and build dashboards, Library gained file and folder sharing, and Box, Dropbox, and SharePoint joined Google Drive in Library OpenAI launched ChatGPT for Financial Services with built-in premium data from Daloopa, PitchBook, LSEG News, and Crunchbase, shaped with Morgan Stanley and Evercore, and a GSA agreement gives US governments $0 license fees, 50% off usage, and Daybreak Blue at half price Smaller ChatGPT bits: a stock watchlist in Finances for US Plus and Pro, a small business plugin collection, over 5 million ChatGPT Sites built in three months, and the desktop pet can now start a new chat with a new Mini option OpenAI published The Work Now Within Reach, calling free access "supported by advertising", citing over one billion weekly active users, and saying they plan to begin deploying their Jalapeño inference chip by year-end OpenAI moved GPT-Rosalind out of research preview for eligible organizations worldwide, detailed Habitat with the service rewritten in Rust by two engineers with Codex and the platform serving over 70 million requests per second, and shared a case study of GPT-5.6 Sol calibrating a six-qubit chip at MIT Paul Christiano joined the OpenAI Foundation Board and the Safety and Security Committee, OpenAI committed $5 million to research on AI and teens, and expanded journalism programs from CUNY and Northwestern to Ukrainian newsrooms Anthropic published an alignment assessment of four incidents in which Claude models gained unauthorized access to real systems, adding a fourth found when assembling transcripts for METR, walking back their July 30 claims, and calling it a mistake that Claude Mythos 5 shipped without alignment environments Anthropic's most detailed threat intelligence report yet says Moonshot and DeepSeek silently relayed their own users' requests to Claude and served the responses as Kimi and DeepSeek output, with distillation campaigns attributed to Alibaba as the largest ever with more than 3,500 fraudulent accounts, plus Zhipu, Xiaomi, SenseTime, and MiniMax Anthropic's Frontier Red Team measured tactical intelligence targeting and conventional weapons capabilities, with Claude Mythos Preview leading the targeting evals, Claude Opus 5 leading the weapons software evals, and Mythos beating the top GeoGuessr division on photo geolocation Anthropic CEO Dario Amodei argues in We Must Pace the Frontier that the AI industry should slow down, commits to giving evaluators like METR permanent employee-level access, and OpenAI CEO Sam Altman replied that he agrees and OpenAI will commit to the same Anthropic's Economics team released a scenario explorer for AI's effect on the US economy by 2030, where even the extreme case grows the economy but reaches 15% annual GDP growth with unemployment beyond recessionary levels On the product side, Claude Code desktop can pop out any pane into its own window, Claude Managed Agents got a session viewer and auto mode, claude plugin eval scores your plugin or skill with and without it, and smart reports launched in beta for Claude Enterprise Anthropic shared lessons from their Claude SMB Tour with more than 1,000 small business owners, where data security was the most-cited adoption barrier and nearly two-thirds asked for more hands-on implementation help, and more

OpenAI and Anthropic this week: Navier-Stokes, An Alien Mind, Images 2.5, Pro pause, Pace the Frontier (Week 37, 2026) OpenAI shared a solution to the Navier-Stokes Millennium Prize Problem, produced in 88 hours by around 10,000 coordinating agents on an internal model still in training and significantly more capable than GPT-6 Astra, with an investigation finding Tristan Buckmaster's Codex prompts could not have influenced the system and no user data was accessed OpenAI says they have reached their automated research intern goal and are making strong progress toward an automated AI researcher by March 2028, with the research org at 3.1 agent-workdays per human workday and RL training on deployment-bound models partly paused after the Hugging Face incident OpenAI Chief Scientist Jakub Pachocki writes in An Alien Mind that no lab has solved alignment and monitoring well enough to keep scaling at maximum speed for much longer, and that GPT-6 Astra is the first model to benefit from some of their newer alignment work OpenAI called for mandatory capability-based national AI safety regulation, endorsed four California bills, said fully autonomous recursive self-improvement should not be pursued until it can be done safely, and described Astra safeguards like universal monitoring of full trajectories including chains of thought OpenAI released ChatGPT Images 2.5 to all ChatGPT, ChatGPT Work, and Codex users with sharper details, up to 50% lower latency, comment-based edits, Sketch, and templates, plus GPT-Image-2.5 Flare and Sunburst in the API ChatGPT Voice can now use GPT-5.6 Sol and GPT-6 Astra when it needs to search or reason, GPT-Live-1 daily limits are simplified per plan, and extra Voice usage drops from 5 to 1.25 credits per minute for Business and Enterprise workspaces on credits OpenAI paused new subscriptions to the $200 ChatGPT Pro plan to protect access for existing GPT-6 Astra users, hours after a remote switch to pause Pro 20x purchases showed up in the web app Custom GPTs in ChatGPT will likely be retired on December 11 based on my findings, and OpenAI's new FAQ puts Enterprise migration at September 17 with new GPT creation ending September 25 and instructions becoming a plugin skill From the ChatGPT Android build, OpenAI is building a collaborative multiplayer document editor with dedicated gateway hosts and draft-conflict UI, an Artifacts Library with Favorites, and a credit score feature in ChatGPT Finance, and Locked Chats with a PIN are being prepared in the web app OpenAI released GPT-Live-1 in the API at $0.05 per minute, a voice model that listens and speaks at the same time and delegates reasoning to a backend model like GPT-6 Astra, and the Agents API in public beta running the Codex harness on OpenAI's infrastructure with only the sandbox left for you to choose Deep research arrived in ChatGPT Work and Codex, a Data plugin connects Snowflake, Databricks, BigQuery, and Redshift to answer business questions and build dashboards, Library gained file and folder sharing, and Box, Dropbox, and SharePoint joined Google Drive in Library OpenAI launched ChatGPT for Financial Services with built-in premium data from Daloopa, PitchBook, LSEG News, and Crunchbase, shaped with Morgan Stanley and Evercore, and a GSA agreement gives US governments $0 license fees, 50% off usage, and Daybreak Blue at half price Smaller ChatGPT bits: a stock watchlist in Finances for US Plus and Pro, a small business plugin collection, over 5 million ChatGPT Sites built in three months, and the desktop pet can now start a new chat with a new Mini option OpenAI published The Work Now Within Reach, calling free access "supported by advertising", citing over one billion weekly active users, and saying they plan to begin deploying their Jalapeño inference chip by year-end OpenAI moved GPT-Rosalind out of research preview for eligible organizations worldwide, detailed Habitat with the service rewritten in Rust by two engineers with Codex and the platform serving over 70 million requests per second, and shared a case study of GPT-5.6 Sol calibrating a six-qubit chip at MIT Paul Christiano joined the OpenAI Foundation Board and the Safety and Security Committee, OpenAI committed $5 million to research on AI and teens, and expanded journalism programs from CUNY and Northwestern to Ukrainian newsrooms Anthropic published an alignment assessment of four incidents in which Claude models gained unauthorized access to real systems, adding a fourth found when assembling transcripts for METR, walking back their July 30 claims, and calling it a mistake that Claude Mythos 5 shipped without alignment environments Anthropic's most detailed threat intelligence report yet says Moonshot and DeepSeek silently relayed their own users' requests to Claude and served the responses as Kimi and DeepSeek output, with distillation campaigns attributed to Alibaba as the largest ever with more than 3,500 fraudulent accounts, plus Zhipu, Xiaomi, SenseTime, and MiniMax Anthropic's Frontier Red Team measured tactical intelligence targeting and conventional weapons capabilities, with Claude Mythos Preview leading the targeting evals, Claude Opus 5 leading the weapons software evals, and Mythos beating the top GeoGuessr division on photo geolocation Anthropic CEO Dario Amodei argues in We Must Pace the Frontier that the AI industry should slow down, commits to giving evaluators like METR permanent employee-level access, and OpenAI CEO Sam Altman replied that he agrees and OpenAI will commit to the same Anthropic's Economics team released a scenario explorer for AI's effect on the US economy by 2030, where even the extreme case grows the economy but reaches 15% annual GDP growth with unemployment beyond recessionary levels On the product side, Claude Code desktop can pop out any pane into its own window, Claude Managed Agents got a session viewer and auto mode, claude plugin eval scores your plugin or skill with and without it, and smart reports launched in beta for Claude Enterprise Anthropic shared lessons from their Claude SMB Tour with more than 1,000 small business owners, where data security was the most-cited adoption barrier and nearly two-thirds asked for more hands-on implementation help, and more

Tibor Blaho

14,925 views • 8 days ago

ChatGPT 6 Astra who? 😭 Fable 5.1 is mogging.. in blender - u just asked it to make 1km x 1km WOW-like fantasy world in blender - basically, just gave it the theme: two opposing fantastical forces - Fable took care of terrains, buildings, trees, points of interest, assets, and layout of the scene - used Blender MCP + Blender skills - Even plugged into local comfyUI setup for texture generation - Then it created the aerial views + 30-60 second flythrough - all from a single prompt.. -This took about 20-25 minutes / 400k tokens :D What is really amazing is how far prompt + tool/MCP + engineering has come now you’re not just asking the AI to make an image anymore.. You’re just asking it to just access the tools and telling it: "hey bro build this game..." And it does... If that’s what a one shot test looks like, game developers are cooked...

ChatGPT 6 Astra who? 😭 Fable 5.1 is mogging.. in blender - u just asked it to make 1km x 1km WOW-like fantasy world in blender - basically, just gave it the theme: two opposing fantastical forces - Fable took care of terrains, buildings, trees, points of interest, assets, and layout of the scene - used Blender MCP + Blender skills - Even plugged into local comfyUI setup for texture generation - Then it created the aerial views + 30-60 second flythrough - all from a single prompt.. -This took about 20-25 minutes / 400k tokens :D What is really amazing is how far prompt + tool/MCP + engineering has come now you’re not just asking the AI to make an image anymore.. You’re just asking it to just access the tools and telling it: "hey bro build this game..." And it does... If that’s what a one shot test looks like, game developers are cooked...

Priy

51,025 views • 17 days ago

Run GameTerminal for a single session, describe the game you want to build, and watch it come together. No game engine to learn. No external tools to switch between. No endless setup before you can start creating. Just describe your idea. GameTerminal plans the architecture, generates the gameplay logic, builds the world, creates assets, writes the code, iterates on the output, and turns your prompt into a playable experience. Whether it's a browser game, a 3D experience built with modern web technologies, or an on-chain game with blockchain interactions, everything happens through one creation workflow. Need physics? Character animations? UI? Game mechanics? Shaders? Optimization? Our AI engine coordinates the entire pipeline so creators can focus on ideas instead of infrastructure. And when you're ready to go beyond traditional games, GameTerminal lets you create blockchain-native experiences with on-chain transactions, digital assets, token economies, and wallet integration—without requiring deep Web3 expertise. This isn't about replacing developers. It's about giving every creator the ability to build at a speed that wasn't possible before. What once took a small studio weeks or months can now begin with a single prompt and evolve in minutes. We're still at the beginning. The way games are created is changing. We believe the next generation of games won't start with code. They'll start with an idea. Describe it. We'll build it.

Run GameTerminal for a single session, describe the game you want to build, and watch it come together. No game engine to learn. No external tools to switch between. No endless setup before you can start creating. Just describe your idea. GameTerminal plans the architecture, generates the gameplay logic, builds the world, creates assets, writes the code, iterates on the output, and turns your prompt into a playable experience. Whether it's a browser game, a 3D experience built with modern web technologies, or an on-chain game with blockchain interactions, everything happens through one creation workflow. Need physics? Character animations? UI? Game mechanics? Shaders? Optimization? Our AI engine coordinates the entire pipeline so creators can focus on ideas instead of infrastructure. And when you're ready to go beyond traditional games, GameTerminal lets you create blockchain-native experiences with on-chain transactions, digital assets, token economies, and wallet integration—without requiring deep Web3 expertise. This isn't about replacing developers. It's about giving every creator the ability to build at a speed that wasn't possible before. What once took a small studio weeks or months can now begin with a single prompt and evolve in minutes. We're still at the beginning. The way games are created is changing. We believe the next generation of games won't start with code. They'll start with an idea. Describe it. We'll build it.

Game Terminal

12,637 views • 1 month ago

THIS GUY VIBE CODED A FULL CAPYBARA FOOD DELIVERY GAME IN 2 WEEKS WITH CLAUDE CODE you play as a capybara delivering food on a bike. orders stack on the back, you have a phone with apps in-game and the whole delivery system is realistic 2 weeks, zero game dev experience, and ENTIRELY AI generated the full stack: > claude code for all the code > three.js for the 3D engine > suno for original music > elevenlabs for sound effects and voice > GPT images-2 and grok for textures and illustrations > tripo3d for generating all the 3D assets the cinematics are all in-game too. he asked claude to build a cinematic editor with timeline controls, camera animation, and transitions. then he just placed the cameras himself his workflow was more planning than coding (obviously): > come up with the core mechanic > plan every feature using claude /plan mode > generate assets with AI tools > spend most of his time on the final polish, prop placement, and making the design feel right he said the human part is what most vibe coded games are missing. AI can generate everything but having taste for what looks good and what feels right is still on you the game is playable right now in the browser this is what vibe coding is actually capable of in the game dev space right now a year ago this would have taken a small team of developers, a sound designer, and an artist working together for months now one person with no experience can ship a polished playable game with story, music, and mechanics in 14 days the tools keep getting better and the barrier to making real games keeps getting lower

THIS GUY VIBE CODED A FULL CAPYBARA FOOD DELIVERY GAME IN 2 WEEKS WITH CLAUDE CODE you play as a capybara delivering food on a bike. orders stack on the back, you have a phone with apps in-game and the whole delivery system is realistic 2 weeks, zero game dev experience, and ENTIRELY AI generated the full stack: > claude code for all the code > three.js for the 3D engine > suno for original music > elevenlabs for sound effects and voice > GPT images-2 and grok for textures and illustrations > tripo3d for generating all the 3D assets the cinematics are all in-game too. he asked claude to build a cinematic editor with timeline controls, camera animation, and transitions. then he just placed the cameras himself his workflow was more planning than coding (obviously): > come up with the core mechanic > plan every feature using claude /plan mode > generate assets with AI tools > spend most of his time on the final polish, prop placement, and making the design feel right he said the human part is what most vibe coded games are missing. AI can generate everything but having taste for what looks good and what feels right is still on you the game is playable right now in the browser this is what vibe coding is actually capable of in the game dev space right now a year ago this would have taken a small team of developers, a sound designer, and an artist working together for months now one person with no experience can ship a polished playable game with story, music, and mechanics in 14 days the tools keep getting better and the barrier to making real games keeps getting lower

Om Patel

241,709 views • 4 months ago

🚨 We’re officially launching : ANIMA STUDIO 🚨 ⭐️ To celebrate, we’re kicking off a 30-DAY CHALLENGE ⭐️ Every day, you’ll find a new case study across our platforms – showcasing our creative vision and technical expertise. Expect: - Brand ads (real or fictional collabs) 👟 - Short films 🎥 - Product reels 📱 - Character design 🦖 - Set design ⛰️ - Art direction 🎨 ..and more. All powered by AI tools + design expertise. We've been producing visuals (images & videos) for brands, entrepreneurs & companies for over 6 months — and have been doing it manually for over 3 years. But Anima Studio is a much bigger vision. We’re blending our roots in design with cutting-edge AI to build something new. In fact, months before this announcement, we started working with a major US production (multi-million budget). 🇺🇸 We can’t say more for now — but it’s moving fast. Our studio covers a full creative scope: - Art Direction 🎨 - Visual / Communication Design & Branding 👁️ - Ad Creation & Deep Brand Identity 🛜 And for Film & Entertainment 🎬 : - Pre-production (worldbuilding, chara & set design) ⭐️ - Production/Post-prod (AI animation, VFX…) 📺 - Transmedia (comics, merch, narrative extensions…) 🪐 - Short Film 🍿 And a lot of innovative ideas to come. We’re still continuing our content creation & training activities on AI & design. Today, we’re just sharing a fun little teaser to launch the challenge — but a proper trailer is coming soon. 💛 Thank you all for the support. 🙏 This wouldn’t be possible without your encouragement. See you tomorrow for the first fictional ad of the challenge! Let’s go. 🚀

Anima

75,478 views • 1 year ago

Everyone's talking, reading, or writing about how much is changing in how we build products in the future, but very few people have actually experienced what the most cutting edge tools are capable of. In a new series on my podcast, I'm going to sit down with the founders of the most advanced product development tools to live demo what these tools can do, and then talk about the implications of this on product people's careers and lives. To kick things off, I sat down with Amjad Masad (Amjad Masad), co-founder and CEO of Replit ⠕, a browser-based coding environment that allows anyone to write and deploy code. Replit has 34 million users globally and is one of the fastest-growing developer communities in the world. In our conversation, Amjad shares: 🔸 A live demo of Replit building a full-stack web app from a text prompt 🔸 The implications of AI-powered development for product managers, designers, and engineers 🔸 What skills will matter more, and less, in the future 🔸 How this might reshape companies and careers 🔸 Why being “generative” will become an increasingly valuable skill 🔸 “Amjad’s law” and how learning to debug AI-generated code is becoming ever more valuable 🔸 Much more Listen now 👇 - YouTube: - Spotify: - Apple: Thank you to our wonderful sponsors for supporting the podcast: 🏆 WorkOS — Modern identity platform for B2B SaaS, free up to 1 million MAUs: 🏆 — A global leader in digital identity verification: 🏆 @LinkedInMktg — Reach professionals and drive results for your business:

Everyone's talking, reading, or writing about how much is changing in how we build products in the future, but very few people have actually experienced what the most cutting edge tools are capable of. In a new series on my podcast, I'm going to sit down with the founders of the most advanced product development tools to live demo what these tools can do, and then talk about the implications of this on product people's careers and lives. To kick things off, I sat down with Amjad Masad (Amjad Masad), co-founder and CEO of Replit ⠕, a browser-based coding environment that allows anyone to write and deploy code. Replit has 34 million users globally and is one of the fastest-growing developer communities in the world. In our conversation, Amjad shares: 🔸 A live demo of Replit building a full-stack web app from a text prompt 🔸 The implications of AI-powered development for product managers, designers, and engineers 🔸 What skills will matter more, and less, in the future 🔸 How this might reshape companies and careers 🔸 Why being “generative” will become an increasingly valuable skill 🔸 “Amjad’s law” and how learning to debug AI-generated code is becoming ever more valuable 🔸 Much more Listen now 👇 - YouTube: - Spotify: - Apple: Thank you to our wonderful sponsors for supporting the podcast: 🏆 WorkOS — Modern identity platform for B2B SaaS, free up to 1 million MAUs: 🏆 — A global leader in digital identity verification: 🏆 @LinkedInMktg — Reach professionals and drive results for your business:

Lenny Rachitsky

81,990 views • 1 year ago

🚨 CHINA'S AI PLAYBOOK ISN'T TO BEAT OPENAI AND ANTHROPIC. IT'S TO MAKE THEM IRRELEVANT BY MAKING INTELLIGENCE FREE. Chinese labs are giving away frontier-level models for free. Zhipu's GLM-5 already beats Claude Opus 4.5 and Gemini 3.0 Pro on key benchmarks. DeepSeek, Qwen, and Kimi round out 4 of the top 5 open-weight models globally, all released under open licenses anyone can use commercially. The real weapon isn't the models, it's the cost to run them. Chinese provinces like Gansu, Guizhou, and Inner Mongolia are cutting data center power bills by up to 50%. DeepSeek trained its R1 model for around $6 million, compared to roughly $100 million for GPT-4. The remaining blocker is chips. US export controls still cut China off from Nvidia's best hardware. But Zhipu already trained a full frontier model entirely on Huawei's Ascend chips, no Nvidia involved. Huawei's next-gen Atlas 950 cluster, launching late 2026, claims 6.7 times the compute of Nvidia's planned NVL144 systems. If Huawei closes the chip gap while China keeps giving away free, near frontier models running on subsidized power. The entire business model OpenAI and Anthropic are built on, charging hundreds of billions in aggregate revenue for access to intelligence, stops making sense. Why pay premium prices for marginal capability gains when a free alternative running on cheap compute gets you 90% of the way there.

🚨 CHINA'S AI PLAYBOOK ISN'T TO BEAT OPENAI AND ANTHROPIC. IT'S TO MAKE THEM IRRELEVANT BY MAKING INTELLIGENCE FREE. Chinese labs are giving away frontier-level models for free. Zhipu's GLM-5 already beats Claude Opus 4.5 and Gemini 3.0 Pro on key benchmarks. DeepSeek, Qwen, and Kimi round out 4 of the top 5 open-weight models globally, all released under open licenses anyone can use commercially. The real weapon isn't the models, it's the cost to run them. Chinese provinces like Gansu, Guizhou, and Inner Mongolia are cutting data center power bills by up to 50%. DeepSeek trained its R1 model for around $6 million, compared to roughly $100 million for GPT-4. The remaining blocker is chips. US export controls still cut China off from Nvidia's best hardware. But Zhipu already trained a full frontier model entirely on Huawei's Ascend chips, no Nvidia involved. Huawei's next-gen Atlas 950 cluster, launching late 2026, claims 6.7 times the compute of Nvidia's planned NVL144 systems. If Huawei closes the chip gap while China keeps giving away free, near frontier models running on subsidized power. The entire business model OpenAI and Anthropic are built on, charging hundreds of billions in aggregate revenue for access to intelligence, stops making sense. Why pay premium prices for marginal capability gains when a free alternative running on cheap compute gets you 90% of the way there.

The Macro Paper

71,921 views • 2 months ago