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.

A parametric drone model in nTop generated 100+ simulation-ready configs, zero geometry failures. PhysicsX.ai ran CFD on all of them, trained a surrogate within ~5%, then ran it live inside nTop to predict pressure fields in under a second as geometry changes. AI-accelerated design exploration with nTop + PhysicsX... show more

nTop

6,018 subscribers

44,419 views • 1 month ago •via X (Twitter)

Education News & Politics Science & Technology

Anya Rossi• Live Now

Private livecam show

0 Comments

No comments available

Comments from the original post will appear here

Related Videos

Since he’s not in Paris anymore, here it is: Hudson Williams yesterday was at the balcony with Aika Flores (his hairstylist and friend) signing autographs and sending them as little paper planes and then he shouted that he ran out of paper for them to be careful with the heads and threw a clementine. via sofianealarcon & _anais_louise

Since he’s not in Paris anymore, here it is: Hudson Williams yesterday was at the balcony with Aika Flores (his hairstylist and friend) signing autographs and sending them as little paper planes and then he shouted that he ran out of paper for them to be careful with the heads and threw a clementine. via sofianealarcon & _anais_louise

Hudson Williams Updates

289,796 views • 4 months ago

Before the part ever touches the machine, it's already been built. This is a 5-axis toolpath, programmed in CAM to map every angle, every pass and every cut. When it's ready, that program goes straight to our DMG MORI. The machine rotates and the tool engages. Then, it executes exactly what you see on screen, with extreme precision. Five axes of motion provide complex geometry in a single setup, with better accuracy. This is precision manufacturing for satellites at our factory in Northern California. 🇺🇸

Before the part ever touches the machine, it's already been built. This is a 5-axis toolpath, programmed in CAM to map every angle, every pass and every cut. When it's ready, that program goes straight to our DMG MORI. The machine rotates and the tool engages. Then, it executes exactly what you see on screen, with extreme precision. Five axes of motion provide complex geometry in a single setup, with better accuracy. This is precision manufacturing for satellites at our factory in Northern California. 🇺🇸

Astranis Space Technologies

12,558 views • 3 months ago

✨ 3d models are now LIVE on Photo AI 😊 You can now turn any AI photo you make into a 3d model by pressing [ 📦 Make 3d model ] And then you can view it inside Photo AI or download it as a .GLB 3d model file It's still very early in AI generated 3d model world but it's nice to have this feature working already As always, the models will keep improving, so this feature will keep getting better (like it did with video, it sucked before, now it's getting passable) Next would be nice to switch to .USDZ so you can load it straight into your iPhone with ARKit and put it in your room Available now for everyone on the Premium and Ultra plan

✨ 3d models are now LIVE on Photo AI 😊 You can now turn any AI photo you make into a 3d model by pressing [ 📦 Make 3d model ] And then you can view it inside Photo AI or download it as a .GLB 3d model file It's still very early in AI generated 3d model world but it's nice to have this feature working already As always, the models will keep improving, so this feature will keep getting better (like it did with video, it sucked before, now it's getting passable) Next would be nice to switch to .USDZ so you can load it straight into your iPhone with ARKit and put it in your room Available now for everyone on the Premium and Ultra plan

@levelsio

112,930 views • 1 year ago

Merkaba isn't the vehicle, it's the road. Pleiadians teach us that "merkaba" is the divine geometric structure permeating our reality which can be harnessed to open portals. People in ancient times confused the opening of portals with the crafts traveling through them. Higher beings don't literally travel in sacred geometry flying in the air. Merkaba is the natural force that will allow the portals to open and shift us to Sheen at Elder Ikai's command. ------- #NEIOH - "In ancient dreams, many large crafts were described as wheels within wheels as lights covered the skies in mesmerizing colors. Many of these shapes were Merkaba and this shape is still noted on Earth as sacred geometry. All life forms contain patterns of sacred geometry. Whether a leaf, tree or rock, a pattern of creation is present. Your body, the water you drink and all creatures that live in your world are composed of holographic patterns of beauty. You are not random in your intricate nature of being. You were planned and created with a masterful intent. As you move and live in this moment, find peace in knowing that one spirit holds all life. You only need to shift your awareness to feel the enormity of gifts available. Know that living Merkaba surround everything in a frequency unseen. Merkaba are extremely powerful and are there to shift you into 5-D Sheen. The great portal around the Earth will transport spirit and consciousness in one instant. Merkaba is the frequency that powers the moment of this change."

Merkaba isn't the vehicle, it's the road. Pleiadians teach us that "merkaba" is the divine geometric structure permeating our reality which can be harnessed to open portals. People in ancient times confused the opening of portals with the crafts traveling through them. Higher beings don't literally travel in sacred geometry flying in the air. Merkaba is the natural force that will allow the portals to open and shift us to Sheen at Elder Ikai's command. ------- #NEIOH - "In ancient dreams, many large crafts were described as wheels within wheels as lights covered the skies in mesmerizing colors. Many of these shapes were Merkaba and this shape is still noted on Earth as sacred geometry. All life forms contain patterns of sacred geometry. Whether a leaf, tree or rock, a pattern of creation is present. Your body, the water you drink and all creatures that live in your world are composed of holographic patterns of beauty. You are not random in your intricate nature of being. You were planned and created with a masterful intent. As you move and live in this moment, find peace in knowing that one spirit holds all life. You only need to shift your awareness to feel the enormity of gifts available. Know that living Merkaba surround everything in a frequency unseen. Merkaba are extremely powerful and are there to shift you into 5-D Sheen. The great portal around the Earth will transport spirit and consciousness in one instant. Merkaba is the frequency that powers the moment of this change."

Kab

95,665 views • 9 months ago

As announced in partnership with NVIDIA at CES, we’re excited to introduce Stable Point Aware 3D (SPAR3D), setting a new standard in 3D generation. Ideal for running on NVIDIA RTX AI PCs, SPAR3D enables real-time editing and complete structure generation of 3D objects from a single image in under a second. You can download the weights on Hugging Face and code on GitHub, or access the model through the Stability AI API. Learn more here: (1/3)

As announced in partnership with NVIDIA at CES, we’re excited to introduce Stable Point Aware 3D (SPAR3D), setting a new standard in 3D generation. Ideal for running on NVIDIA RTX AI PCs, SPAR3D enables real-time editing and complete structure generation of 3D objects from a single image in under a second. You can download the weights on Hugging Face and code on GitHub, or access the model through the Stability AI API. Learn more here: (1/3)

Stability AI

181,479 views • 1 year ago

Comparing AI video results: • Gen-3 • Kling AI 1.5 • Hailuo MiniMax • Luma Dream Machine I used a FLUX1.1 [pro] image as the input. Each model ran the same prompt 3 times, and I chose the best results: "Smooth side tracking shot of a woman jogging at twilight with the city skyline in the background, sharp focus on the woman as she pumps her arms with each stride as she jogs, depth of field, advertising commercial" Running, jogging, and walking are difficult for AI video models, and none of them produced a smooth result, but this is the worst they will be.

Comparing AI video results: • Gen-3 • Kling AI 1.5 • Hailuo MiniMax • Luma Dream Machine I used a FLUX1.1 [pro] image as the input. Each model ran the same prompt 3 times, and I chose the best results: "Smooth side tracking shot of a woman jogging at twilight with the city skyline in the background, sharp focus on the woman as she pumps her arms with each stride as she jogs, depth of field, advertising commercial" Running, jogging, and walking are difficult for AI video models, and none of them produced a smooth result, but this is the worst they will be.

Heather Cooper

57,672 views • 1 year ago

A 22-YEAR-OLD IN CHINA SOLD HER VIRAL AI REELS TO 6 LOCAL BUSINESSES AND CLEARED $11,900 LAST MONTH She never filmed a thing. She tracked 40 faceless channels in one sheet and scored every clip that beat its own channel. One line ran the sort: views over the channel's median. Anything past 30 was a format she could reskin that week. The AI-video look was blowing up on Douyin, so she took the shape and ran it in english first. For $50 a month in apps she found the format on Monday and shipped by Wednesday. Claude prompted each format into 30 shot lists. CapCut generated the whole clip, a photoreal AI girl in a full scene, with AI voice and captions. Make posted to 3 platforms and fed the numbers back into the same sheet, so her own hits fed the next batch. She walked into a boutique with a phone and a page of outlier scores. She left with a retainer for 20 AI videos a month. A studio needs a model, a crew, and a shoot day for one clip. She generated hers overnight. Every prompt and the scenario are in the article.

A 22-YEAR-OLD IN CHINA SOLD HER VIRAL AI REELS TO 6 LOCAL BUSINESSES AND CLEARED $11,900 LAST MONTH She never filmed a thing. She tracked 40 faceless channels in one sheet and scored every clip that beat its own channel. One line ran the sort: views over the channel's median. Anything past 30 was a format she could reskin that week. The AI-video look was blowing up on Douyin, so she took the shape and ran it in english first. For $50 a month in apps she found the format on Monday and shipped by Wednesday. Claude prompted each format into 30 shot lists. CapCut generated the whole clip, a photoreal AI girl in a full scene, with AI voice and captions. Make posted to 3 platforms and fed the numbers back into the same sheet, so her own hits fed the next batch. She walked into a boutique with a phone and a page of outlier scores. She left with a retainer for 20 AI videos a month. A studio needs a model, a crew, and a shoot day for one clip. She generated hers overnight. Every prompt and the scenario are in the article.

Fokki

80,741 views • 16 days ago

The architecture of this new world model is one of the most interesting things I've seen lately: Let me first explain how most world models work: They predict and render one frame at a time. If you are navigating in one of these worlds, and you look left, the model draws whatever looks right in the moment. Every time you change your viewpoint, the model has to imagine what should be there again, so it's very common for these models to "forget" what's in the world. For example, if you put a toy on the table, look away, then look back, the toy might not be there anymore. Tripo AI is releasing its Project Eden model, which works very differently: The model builds the world first, and then renders it based on that map. That map holds the real state of the world: the geometry, every object, where things are, what's already happened. The picture you see on screen gets generated from the map. This architecture flips the whole thing. Now, you get the following: 1. The world stops forgetting. Leave, come back, and the toy is still on the table because it lives in the map, not in the last frame you saw. 2. You can edit the world, and those changes persist for anyone who enters later. 3. Multiple people and AI agents can coexist in the world and see it from different perspectives. This is early research, but it's looking really promising. They just raised nearly $200M across two rounds to build it out. Tripo will be at SIGGRAPH 2026 (July 19–23, Los Angeles Convention Center). If you work in 3D, embodied AI, simulation, or anything spatial, go connect with them there.

The architecture of this new world model is one of the most interesting things I've seen lately: Let me first explain how most world models work: They predict and render one frame at a time. If you are navigating in one of these worlds, and you look left, the model draws whatever looks right in the moment. Every time you change your viewpoint, the model has to imagine what should be there again, so it's very common for these models to "forget" what's in the world. For example, if you put a toy on the table, look away, then look back, the toy might not be there anymore. Tripo AI is releasing its Project Eden model, which works very differently: The model builds the world first, and then renders it based on that map. That map holds the real state of the world: the geometry, every object, where things are, what's already happened. The picture you see on screen gets generated from the map. This architecture flips the whole thing. Now, you get the following: 1. The world stops forgetting. Leave, come back, and the toy is still on the table because it lives in the map, not in the last frame you saw. 2. You can edit the world, and those changes persist for anyone who enters later. 3. Multiple people and AI agents can coexist in the world and see it from different perspectives. This is early research, but it's looking really promising. They just raised nearly $200M across two rounds to build it out. Tripo will be at SIGGRAPH 2026 (July 19–23, Los Angeles Convention Center). If you work in 3D, embodied AI, simulation, or anything spatial, go connect with them there.

Santiago

30,189 views • 1 month ago

🚨 AI MAY HAVE JUST SHATTERED A MATHEMATICAL BELIEF THAT STOOD FOR 80 YEARS. A geometry problem posed by legendary mathematician Paul Erdős in 1946 may have just been overturned… by an AI. The problem sounds simple on the surface: How many pairs of dots can exist on a flat plane while all being exactly the same distance apart? For decades, mathematicians believed there was a strict upper limit to how fast those equal-distance pairs could grow. But OpenAI’s internal model found something unexpected: There may be no universal limit at all. Using high-dimensional geometry and abstract number structures far beyond human visual intuition, the AI generated hundreds of pages of rigorous reasoning showing you can create vastly more unit-distance relationships than experts ever thought possible. Why this matters: It challenges one of the longest-standing open problems in discrete geometry. It shows AI isn’t just solving known problems faster it’s discovering entirely new mathematical territory. It raises the real possibility that machines can now explore abstract realities humans never even thought to search for. The deeper implication is almost unsettling: For centuries, mathematics was considered the purest form of human reasoning. Now AI may be helping expand the very boundaries of proof itself finding truths no human mind can fully grasp intuitively. What happens when machines start discovering mathematical realities that feel alien even to the mathematicians who study them?

🚨 AI MAY HAVE JUST SHATTERED A MATHEMATICAL BELIEF THAT STOOD FOR 80 YEARS. A geometry problem posed by legendary mathematician Paul Erdős in 1946 may have just been overturned… by an AI. The problem sounds simple on the surface: How many pairs of dots can exist on a flat plane while all being exactly the same distance apart? For decades, mathematicians believed there was a strict upper limit to how fast those equal-distance pairs could grow. But OpenAI’s internal model found something unexpected: There may be no universal limit at all. Using high-dimensional geometry and abstract number structures far beyond human visual intuition, the AI generated hundreds of pages of rigorous reasoning showing you can create vastly more unit-distance relationships than experts ever thought possible. Why this matters: It challenges one of the longest-standing open problems in discrete geometry. It shows AI isn’t just solving known problems faster it’s discovering entirely new mathematical territory. It raises the real possibility that machines can now explore abstract realities humans never even thought to search for. The deeper implication is almost unsettling: For centuries, mathematics was considered the purest form of human reasoning. Now AI may be helping expand the very boundaries of proof itself finding truths no human mind can fully grasp intuitively. What happens when machines start discovering mathematical realities that feel alien even to the mathematicians who study them?

TheNewPhysics

15,516 views • 2 months ago

✨ Every week a new AI model comes out and it suddenly makes my half broken features work a lot better Yesterday Seedream-4-Edit came out and it made my [ Hold product ] feature on Photo AI a lot better You can now go from: 🎁 Product photo -> 👱‍♀️ Talking video with your AI model while holding your product. In just a few minutes! Here's a photo I took from the weekly farm box we get in our kitchen, I set it as the product and then with Photo AI made it into a talking video where my trained AI model presents it It's not perfect, as the objects inside the farm box still move around a bit, but pretty close. If the product is more uniform (like lip gloss, a product box or a book) it does a pretty good job at keeping it exactly the same This "consistency" as they call it is quite important for actual real world use. Product sellers don't want to have an image or video of an AI model if the product doesn't look exactly the same as what they sell With that, I'm getting pretty close now and every week with every new model that comes out, a bit closer And it's interesting cause now I'm finally moving from B2C a bit more to B2B where businesses can use Photo AI more, designers and stores already use it for trying on clothes etc. but now they can generate content for real products! 😊 LIVE now on Photo AI

✨ Every week a new AI model comes out and it suddenly makes my half broken features work a lot better Yesterday Seedream-4-Edit came out and it made my [ Hold product ] feature on Photo AI a lot better You can now go from: 🎁 Product photo -> 👱‍♀️ Talking video with your AI model while holding your product. In just a few minutes! Here's a photo I took from the weekly farm box we get in our kitchen, I set it as the product and then with Photo AI made it into a talking video where my trained AI model presents it It's not perfect, as the objects inside the farm box still move around a bit, but pretty close. If the product is more uniform (like lip gloss, a product box or a book) it does a pretty good job at keeping it exactly the same This "consistency" as they call it is quite important for actual real world use. Product sellers don't want to have an image or video of an AI model if the product doesn't look exactly the same as what they sell With that, I'm getting pretty close now and every week with every new model that comes out, a bit closer And it's interesting cause now I'm finally moving from B2C a bit more to B2B where businesses can use Photo AI more, designers and stores already use it for trying on clothes etc. but now they can generate content for real products! 😊 LIVE now on Photo AI

@levelsio

361,558 views • 10 months ago

We just launched Claude Sonnet 4.5 and “Imagine with Claude”, a 5-day preview that creates dynamic experiences for you in real-time (and the first product I’ve worked on at Anthropic!). It pioneers the concept of “model-as-backend”, using a model to not only generate interfaces on the fly but also power all the functionality behind it – all made possible by our newest model, Claude Sonnet 4.5. Here’s a "choose your own adventure" version of my founder journey, as retold by Imagine with Claude:

We just launched Claude Sonnet 4.5 and “Imagine with Claude”, a 5-day preview that creates dynamic experiences for you in real-time (and the first product I’ve worked on at Anthropic!). It pioneers the concept of “model-as-backend”, using a model to not only generate interfaces on the fly but also power all the functionality behind it – all made possible by our newest model, Claude Sonnet 4.5. Here’s a "choose your own adventure" version of my founder journey, as retold by Imagine with Claude:

Sean Strong

115,693 views • 9 months ago

If this contract had gone live, attackers would have drained every single ETH within hours. A crowdfunding platform. Clean-looking Solidity. Ready for mainnet. The dev team was confident. Then we ran it through V2's AI Smart Contract Auditor. 9 vulnerabilities detected. 1 critical. 2 high. Under 20 minutes. The critical finding? A textbook reentrancy bug in claimRefund(). The contract sent ETH before zeroing the user's balance. An attacker could re-enter the function during the transfer and drain the same refund repeatedly until the entire contract was empty. Every backer's funds across every campaign, gone in one transaction. It also caught an emergencyWithdraw() function that let the owner pull the entire contract balance with zero restrictions. Active campaigns, pending refunds, everything. One compromised key and it's a complete rug. Plus an uncapped platform fee that could be set to 100% (creators get nothing) or above 100% (permanent DoS on all fund claims). The AI generated a 24-page report with severity classifications, exact code locations, detailed attack scenarios, and specific fixes for each issue. Final score: 89/100. Not because it was "pretty good," but because the architecture was sound and the issues were fixable. Without catching them though? Catastrophic. A traditional audit runs $5K to $50K and takes weeks. This took under 20 minutes. That's the power of V2 ⚡️

If this contract had gone live, attackers would have drained every single ETH within hours. A crowdfunding platform. Clean-looking Solidity. Ready for mainnet. The dev team was confident. Then we ran it through V2's AI Smart Contract Auditor. 9 vulnerabilities detected. 1 critical. 2 high. Under 20 minutes. The critical finding? A textbook reentrancy bug in claimRefund(). The contract sent ETH before zeroing the user's balance. An attacker could re-enter the function during the transfer and drain the same refund repeatedly until the entire contract was empty. Every backer's funds across every campaign, gone in one transaction. It also caught an emergencyWithdraw() function that let the owner pull the entire contract balance with zero restrictions. Active campaigns, pending refunds, everything. One compromised key and it's a complete rug. Plus an uncapped platform fee that could be set to 100% (creators get nothing) or above 100% (permanent DoS on all fund claims). The AI generated a 24-page report with severity classifications, exact code locations, detailed attack scenarios, and specific fixes for each issue. Final score: 89/100. Not because it was "pretty good," but because the architecture was sound and the issues were fixable. Without catching them though? Catastrophic. A traditional audit runs $5K to $50K and takes weeks. This took under 20 minutes. That's the power of V2 ⚡️

ChainGPT

26,504 views • 4 months ago

Polymorphic moulding: a manufacturing method that forms parts using a grid of computer-controlled pins: Each pin can move up or down independently, so together they shape a surface that acts like a custom mould. Instead of building a fixed mould for every product, the machine quickly repositions the pins to match a new digital model. This means a mould can be created in minutes, used, and then reshaped again for the next design. Because the hardware stays the same and only its configuration changes: > no permanent tooling is needed > very little material is wasted > setup time between products is minimal The system essentially turns mould making into a programmable process. Engineers send a design file, the pins form the mould geometry, and the material is cast or formed inside it. Especially useful for rapid prototyping and customised manufacturing, where many different shapes must be produced in small quantities without rebuilding tools each time. Credit: ---- Weekly robotics and AI insights. Subscribe free:

Polymorphic moulding: a manufacturing method that forms parts using a grid of computer-controlled pins: Each pin can move up or down independently, so together they shape a surface that acts like a custom mould. Instead of building a fixed mould for every product, the machine quickly repositions the pins to match a new digital model. This means a mould can be created in minutes, used, and then reshaped again for the next design. Because the hardware stays the same and only its configuration changes: > no permanent tooling is needed > very little material is wasted > setup time between products is minimal The system essentially turns mould making into a programmable process. Engineers send a design file, the pins form the mould geometry, and the material is cast or formed inside it. Especially useful for rapid prototyping and customised manufacturing, where many different shapes must be produced in small quantities without rebuilding tools each time. Credit: ---- Weekly robotics and AI insights. Subscribe free:

Ilir Aliu

38,234 views • 5 months ago

A 23-year-old girl built a desktop pet and ran a local LLM on a stock 1998 iMac G3 with 32 MB of RAM and Mac OS 8. She didn’t buy any modern upgrades. She used vintage hardware, a cross-compiler, and Claude to bring AI to a 28-year-old machine. Then she ported Karpathy’s llama2.c to classic Mac OS, compiled it with Retro68, used a Python script to swap the model’s data to big-endian for the PowerPC chip, and successfully generated local text via a 260K parameter model. One vintage Mac. 32 megabytes of RAM. Local inference. A free cross-compiler. Most people see old computers as electronic waste. She treated a 233 MHz processor like a live hardware challenge and forced modern AI architecture to run where it was never meant to exist.

A 23-year-old girl built a desktop pet and ran a local LLM on a stock 1998 iMac G3 with 32 MB of RAM and Mac OS 8. She didn’t buy any modern upgrades. She used vintage hardware, a cross-compiler, and Claude to bring AI to a 28-year-old machine. Then she ported Karpathy’s llama2.c to classic Mac OS, compiled it with Retro68, used a Python script to swap the model’s data to big-endian for the PowerPC chip, and successfully generated local text via a 260K parameter model. One vintage Mac. 32 megabytes of RAM. Local inference. A free cross-compiler. Most people see old computers as electronic waste. She treated a 233 MHz processor like a live hardware challenge and forced modern AI architecture to run where it was never meant to exist.

Ridark

35,186 views • 2 months ago

✨ I've brought back from the dead It was my first visual AI project in 2022, and it's this project that generated random ArchDaily-style architecture designs that made me realize AI image models could do interior design That led me to make which then led me to finetune my first interior design model which then for fun I uploaded my own photos too, which led me to make and then pivoted that in to So this project has a special place for me It was still alive but wasn't generating new designs anymore because it ran on Stable Diffusion 1.5 which was outdated and everything stopped working about a year ago I've now migrated it to its own Hetzner VPS now, which means I can run Claude Code on the server with it, and cleaned it up and upgraded it to the latest AI image models (including Nano Banana Pro) It now generates about 12 new designs every day again, and you can up or downvote the ones you like or don't like!

✨ I've brought back from the dead It was my first visual AI project in 2022, and it's this project that generated random ArchDaily-style architecture designs that made me realize AI image models could do interior design That led me to make which then led me to finetune my first interior design model which then for fun I uploaded my own photos too, which led me to make and then pivoted that in to So this project has a special place for me It was still alive but wasn't generating new designs anymore because it ran on Stable Diffusion 1.5 which was outdated and everything stopped working about a year ago I've now migrated it to its own Hetzner VPS now, which means I can run Claude Code on the server with it, and cleaned it up and upgraded it to the latest AI image models (including Nano Banana Pro) It now generates about 12 new designs every day again, and you can up or downvote the ones you like or don't like!

@levelsio

204,387 views • 4 months ago

You can now orchestrate Fable 5, Sol, and any model inside Codex with one plugin. It's called Codex-Orchestration. Assign Fable 5 as the advisor, Sol as the executor, or any model to any role. Then define the order they work in. Codex handles the routing. I ran Fable 5 High as planner with GPT-5.6 Sol Extra High as executor on a set of issues Opus and GPT-5.5 always struggled with. Done in 30 minutes. 40% fewer limit hits. 2x faster implementation. Install it by pasting this into Codex: "Install Codex Orchestration: codex plugin marketplace add Cjbuilds/Codex-Orchestration codex plugin add codex-orchestration@codex-orchestration Verify the installation, then tell me to start a new task." Then assign your models: @ codex-orchestration advisor: Claude Fable 5 High, Executor: GPT-5.6 Sol High Open source. Tweak the routing however you want.

You can now orchestrate Fable 5, Sol, and any model inside Codex with one plugin. It's called Codex-Orchestration. Assign Fable 5 as the advisor, Sol as the executor, or any model to any role. Then define the order they work in. Codex handles the routing. I ran Fable 5 High as planner with GPT-5.6 Sol Extra High as executor on a set of issues Opus and GPT-5.5 always struggled with. Done in 30 minutes. 40% fewer limit hits. 2x faster implementation. Install it by pasting this into Codex: "Install Codex Orchestration: codex plugin marketplace add Cjbuilds/Codex-Orchestration codex plugin add codex-orchestration@codex-orchestration Verify the installation, then tell me to start a new task." Then assign your models: @ codex-orchestration advisor: Claude Fable 5 High, Executor: GPT-5.6 Sol High Open source. Tweak the routing however you want.

Alvaro Cintas

89,540 views • 8 days ago

Introducing Replit ModelFarm, the fastest and safest way to build your next Generative AI app. Available for free on Hacker and Pro plans till October 15th. It requires zero setup, zero configuration, and zero API keys. With Replit ModelFarm, you can build a working Gen AI app in as little as 3 lines of code. Get started by installing the Replit AI library in any Python, JavaScript, or TypeScript Repl. The library implements an API for text completion, chat completion, and text embeddings. It supports streaming so your users can see model responses in real-time rather than waiting on a single output. All Hacker and Pro builders will have free access to a selection of Gen AI models offered by Google Cloud Vertex AI through Replit ModelFarm. All models are accessible from the development environment and any deployed app.

Introducing Replit ModelFarm, the fastest and safest way to build your next Generative AI app. Available for free on Hacker and Pro plans till October 15th. It requires zero setup, zero configuration, and zero API keys. With Replit ModelFarm, you can build a working Gen AI app in as little as 3 lines of code. Get started by installing the Replit AI library in any Python, JavaScript, or TypeScript Repl. The library implements an API for text completion, chat completion, and text embeddings. It supports streaming so your users can see model responses in real-time rather than waiting on a single output. All Hacker and Pro builders will have free access to a selection of Gen AI models offered by Google Cloud Vertex AI through Replit ModelFarm. All models are accessible from the development environment and any deployed app.

Replit ⠕

229,285 views • 2 years ago

This is the most hilarious thing I saw and did today Ran gemma-4-12B-coder-fable5-composer2.5-v1-GGUF locally with 8 GB VRAM at 20+ tok/sec Anthropic's Claude Fable 5 launched June 9. By June 12 it was banned. I can't access it. You can't either. But here's the twist: I'm running a model trained on its chain of thought at 20 tok/s on my RTX 4060 8GB. Locally. Offline. No cloud. No export control. Enter: Gemma4-12B-Coder GGUF (Q4_K_M) Base: Google's gemma-4-12B-it Fine-tuned on verifiable Python CoT data: - Primary: Composer 2.5 real reasoning traces (only passing solutions kept) - Auxiliary: Fable 5 used to redo the hard cases Composer missed. Every training example's reasoning led to code that actually ran. No hallucinated logic. Llama.cpp flags: -m gemma4-coding-Q4_K_M.gguf -cnv -ngl 44 -c 64000 -v (huggingface model link in comments) Flag breakdown: -ngl 44 → offload 44 layers to GPU (tune this for your VRAM) -c 64000 → 64K context window -cnv → conversation/chat mode -v → verbose output The irony writes itself. Anthropic spent weeks telling the world Fable 5 (mythos) is too powerful to release. Then released it. Then got banned from serving it, including their own researchers. Meanwhile: a Gemma 4 12B fine tune, trained on Fable 5's reasoning, runs fully offline on my mid range consumer GPU No API. No cloud. Just me and llama.cpp. This is why local AI matters. Check out the model's link in the comments. How's your experience been with this model?

This is the most hilarious thing I saw and did today Ran gemma-4-12B-coder-fable5-composer2.5-v1-GGUF locally with 8 GB VRAM at 20+ tok/sec Anthropic's Claude Fable 5 launched June 9. By June 12 it was banned. I can't access it. You can't either. But here's the twist: I'm running a model trained on its chain of thought at 20 tok/s on my RTX 4060 8GB. Locally. Offline. No cloud. No export control. Enter: Gemma4-12B-Coder GGUF (Q4_K_M) Base: Google's gemma-4-12B-it Fine-tuned on verifiable Python CoT data: - Primary: Composer 2.5 real reasoning traces (only passing solutions kept) - Auxiliary: Fable 5 used to redo the hard cases Composer missed. Every training example's reasoning led to code that actually ran. No hallucinated logic. Llama.cpp flags: -m gemma4-coding-Q4_K_M.gguf -cnv -ngl 44 -c 64000 -v (huggingface model link in comments) Flag breakdown: -ngl 44 → offload 44 layers to GPU (tune this for your VRAM) -c 64000 → 64K context window -cnv → conversation/chat mode -v → verbose output The irony writes itself. Anthropic spent weeks telling the world Fable 5 (mythos) is too powerful to release. Then released it. Then got banned from serving it, including their own researchers. Meanwhile: a Gemma 4 12B fine tune, trained on Fable 5's reasoning, runs fully offline on my mid range consumer GPU No API. No cloud. Just me and llama.cpp. This is why local AI matters. Check out the model's link in the comments. How's your experience been with this model?

Alok

569,909 views • 1 month ago

Last week, Askari Defense made a large first delivery to an elite unit inside the DoW for a live range demo. After numerous kinetic intercepts against simulated enemy quad-rotors, they were asked about fixed wing intercept capability. As soon as they got back, they generated a CAD model of a fixed-wing drone and imported it file into their FULLY SYNTHETIC vision training pipeline. In other words, they had no real life vision testing data against the fixed wing, but all they needed was the CAD file. 72hrs later, they validated it with a center-mass, fully autonomous intercept. The impact will have on the preservation of human lives is immense.

Last week, Askari Defense made a large first delivery to an elite unit inside the DoW for a live range demo. After numerous kinetic intercepts against simulated enemy quad-rotors, they were asked about fixed wing intercept capability. As soon as they got back, they generated a CAD model of a fixed-wing drone and imported it file into their FULLY SYNTHETIC vision training pipeline. In other words, they had no real life vision testing data against the fixed wing, but all they needed was the CAD file. 72hrs later, they validated it with a center-mass, fully autonomous intercept. The impact will have on the preservation of human lives is immense.

Cam

24,859 views • 1 month ago