Video wird geladen...

Video konnte nicht geladen werden

Beim Laden dieses Videos ist ein Problem aufgetreten. Dies könnte an einem vorübergehenden Netzwerkproblem liegen oder das Video ist möglicherweise nicht verfügbar.

LFM2.5-ColBERT-350M is a surprisingly reliable smart tool selector. We gave it 151 tools, and it consistently surfaces the 5 most relevant ones based on the user prompt. This saves tokens and improves accuracy. Ideal for hmmmm agentic edge models? 👀

Maxime Labonne

33,435 subscribers

19,216 Aufrufe • vor 1 Monat •via X (Twitter)

Anya Rossi• Live Now

Private livecam show

0 Kommentare

Keine Kommentare verfügbar

Kommentare vom Original-Post werden hier angezeigt

Ähnliche Videos

Storing too many tools in your context window increases latency and can lead to wrong tool selection. In this demo, we used LFM2.5-ColBERT-350M as a filter to only select the five most relevant tools among 151 options. It's fast and reliable, even without any specific fine-tuning. Try the demo on Hugging Face! And learn more on our blog:

Storing too many tools in your context window increases latency and can lead to wrong tool selection. In this demo, we used LFM2.5-ColBERT-350M as a filter to only select the five most relevant tools among 151 options. It's fast and reliable, even without any specific fine-tuning. Try the demo on Hugging Face! And learn more on our blog:

Liquid AI

33,677 Aufrufe • vor 1 Monat

A Web-Based, Agentic ComfyUI-Style Builder for 3D Okay, this is actually pretty interesting. I’ve been testing Flow, a new tool from 3DAIStudio . It’s basically a web-based, ComfyUI-style workflow builder made around 3D, with specialized nodes, logic nodes, and an AI agent that can help build the workflow for you. I tested it across a few different pipelines, models, and node setups — and it works surprisingly well. Full review + results in the video 👇

A Web-Based, Agentic ComfyUI-Style Builder for 3D Okay, this is actually pretty interesting. I’ve been testing Flow, a new tool from 3DAIStudio . It’s basically a web-based, ComfyUI-style workflow builder made around 3D, with specialized nodes, logic nodes, and an AI agent that can help build the workflow for you. I tested it across a few different pipelines, models, and node setups — and it works surprisingly well. Full review + results in the video 👇

Stefan 3D AI

10,327 Aufrufe • vor 1 Monat

I cant believe this guy just made a permanent solution to context bloat and open sourced it all! when we tested this tool (Context+) for solving an issue on the OpenCode repository, the agent using this tool used ~6.5k fewer tokens, found the code and fixed it in half the time! the results were surprising: 6 to 10k tokens saved per prompt, completed task in ~2 minutes while the agent running without the tool took ~4 mins for the same and got stuck in loops bro built an entire beast by using all the modern tools that we could think of: undo trees, semantic search by meaning (by haskellforall), advanced refactoring, blast radius, advanced file context trees, restore points... i can keep going on semantic code search and context trees are the future of agentic coding and this tool proves it the feature i loved the most is semantic search and how it gets things done 2x faster with least possible tokens it makes an agent that actually knows what it’s doing and not just guessing, it makes meaning from your code similar to RAG. if you aren't optimizing your context, you are just burning money the developer says this tool is still under development, it can have unexpected behavior and the docs need updates but the video shows the reality of how fast it can be github: get here:

I cant believe this guy just made a permanent solution to context bloat and open sourced it all! when we tested this tool (Context+) for solving an issue on the OpenCode repository, the agent using this tool used ~6.5k fewer tokens, found the code and fixed it in half the time! the results were surprising: 6 to 10k tokens saved per prompt, completed task in ~2 minutes while the agent running without the tool took ~4 mins for the same and got stuck in loops bro built an entire beast by using all the modern tools that we could think of: undo trees, semantic search by meaning (by haskellforall), advanced refactoring, blast radius, advanced file context trees, restore points... i can keep going on semantic code search and context trees are the future of agentic coding and this tool proves it the feature i loved the most is semantic search and how it gets things done 2x faster with least possible tokens it makes an agent that actually knows what it’s doing and not just guessing, it makes meaning from your code similar to RAG. if you aren't optimizing your context, you are just burning money the developer says this tool is still under development, it can have unexpected behavior and the docs need updates but the video shows the reality of how fast it can be github: get here:

forloop

226,054 Aufrufe • vor 5 Monaten

🚀We are excited to introduce the Tool Decathlon (Toolathlon), a benchmark for language agents on diverse, complex, and realistic tool use. ⭐️32 applications and 600+ tools based on real-world software environments ⭐️Execution-based, reliable evaluation ⭐️Realistic, covering daily and professional scenarios Toolathlon reveals significant shortcomings of SOTA LLMs in realistic tool-use tasks, where Claude Sonnet 4.5 achieves 38.6% success rate. It also indicates a clear gap between open-source and leading proprietary models. Check our blog: Github: Paper: 🧵⬇️

🚀We are excited to introduce the Tool Decathlon (Toolathlon), a benchmark for language agents on diverse, complex, and realistic tool use. ⭐️32 applications and 600+ tools based on real-world software environments ⭐️Execution-based, reliable evaluation ⭐️Realistic, covering daily and professional scenarios Toolathlon reveals significant shortcomings of SOTA LLMs in realistic tool-use tasks, where Claude Sonnet 4.5 achieves 38.6% success rate. It also indicates a clear gap between open-source and leading proprietary models. Check our blog: Github: Paper: 🧵⬇️

Junxian He

44,126 Aufrufe • vor 9 Monaten

The MiniMax M2 model is mind-blowing! It's open-source. It outperforms Gemini 2.5, Claude 4.1, and Qwen3 across coding and tool-use benchmarks. Right now, it's one of the world's top 5 models in intelligence! And here is the best part: Claude is one of the best models you can use today, and MiniMax M2 costs only 8% of that! It's smaller, faster, and cheaper. Extremely efficient at using tokens. Minimax M2's biggest strength: High agentic capabilities. The model can plan and execute complex multi-tool workflows. It's reliable and very robust at executing long-horizon tool chains. In summary: • Low latency • Very cheap • Excels at agentic tasks • Open-source The model currently powers the MiniMax Agent and is available for a free global trial. You can access MiniMax M2's API here: To access the agent: And here is the MiniMax website: Thanks to the MiniMax team for showing me the ropes and partnering with me on this post.

The MiniMax M2 model is mind-blowing! It's open-source. It outperforms Gemini 2.5, Claude 4.1, and Qwen3 across coding and tool-use benchmarks. Right now, it's one of the world's top 5 models in intelligence! And here is the best part: Claude is one of the best models you can use today, and MiniMax M2 costs only 8% of that! It's smaller, faster, and cheaper. Extremely efficient at using tokens. Minimax M2's biggest strength: High agentic capabilities. The model can plan and execute complex multi-tool workflows. It's reliable and very robust at executing long-horizon tool chains. In summary: • Low latency • Very cheap • Excels at agentic tasks • Open-source The model currently powers the MiniMax Agent and is available for a free global trial. You can access MiniMax M2's API here: To access the agent: And here is the MiniMax website: Thanks to the MiniMax team for showing me the ropes and partnering with me on this post.

Santiago

91,197 Aufrufe • vor 9 Monaten

Peter Steinberger Quietly Started A Shift That Makes Claude 35x Cheaper To Run The way Claude Code talks to tools matters more than people think. Most use MCP, and it quietly eats tokens, it loads everything into context every time, even tools you never call. One benchmark: MCP burned 35x more tokens than a CLI on the same task. Peter Steinberger, the OpenClaw guy, got annoyed by this and started building lean CLIs himself. That kicked off a tool called Printing Press. You point Claude Code at any website, even ones with no API like ESPN or Craigslist, and it builds a small command tool for it in about 10 minutes. In one demo, a request pulled 132,000 tokens of raw data, but the tool processed it locally and only handed Claude a 2,000-token summary. The rest never touched the context window. It also comes with ~50 ready-made tools you can grab right away. To start, point Claude Code at the links from and ask it to set it up. Bookmark this.

Peter Steinberger Quietly Started A Shift That Makes Claude 35x Cheaper To Run The way Claude Code talks to tools matters more than people think. Most use MCP, and it quietly eats tokens, it loads everything into context every time, even tools you never call. One benchmark: MCP burned 35x more tokens than a CLI on the same task. Peter Steinberger, the OpenClaw guy, got annoyed by this and started building lean CLIs himself. That kicked off a tool called Printing Press. You point Claude Code at any website, even ones with no API like ESPN or Craigslist, and it builds a small command tool for it in about 10 minutes. In one demo, a request pulled 132,000 tokens of raw data, but the tool processed it locally and only handed Claude a 2,000-token summary. The rest never touched the context window. It also comes with ~50 ready-made tools you can grab right away. To start, point Claude Code at the links from and ask it to set it up. Bookmark this.

Ridark

32,351 Aufrufe • vor 1 Monat

Bird SQL is an impressive new tool, based on language models, for searching Twitter. Tools like this are changing the way we interact with information. If used in the right way, signals are becoming easier to find through the use of LLMs. Try it here:

Bird SQL is an impressive new tool, based on language models, for searching Twitter. Tools like this are changing the way we interact with information. If used in the right way, signals are becoming easier to find through the use of LLMs. Try it here:

elvis

95,883 Aufrufe • vor 3 Jahren

Liquid AI's head of post-training explained how they built a small model that runs on-device under 1 GB in 20 minutes - better than $2500 small-model bootcamps. pick LFM2.5 base -> on-policy preference alignment -> agentic reinforcement learning -> curriculum training -> iterative model merging -> ship a 1B model that reliably calls tools on your phone. That loop is why frontier small models now beat 70B models on the tasks that actually matter. LFM2.5 + on-policy DPO + agentic RL + curriculum training + iterative merging - that's the stack. Watch and save it, then run a 1B agent on your phone tonight.

Liquid AI's head of post-training explained how they built a small model that runs on-device under 1 GB in 20 minutes - better than $2500 small-model bootcamps. pick LFM2.5 base -> on-policy preference alignment -> agentic reinforcement learning -> curriculum training -> iterative model merging -> ship a 1B model that reliably calls tools on your phone. That loop is why frontier small models now beat 70B models on the tasks that actually matter. LFM2.5 + on-policy DPO + agentic RL + curriculum training + iterative merging - that's the stack. Watch and save it, then run a 1B agent on your phone tonight.

h100envy

85,953 Aufrufe • vor 17 Tagen

🚨 More DeepSeek V4 Pro GA Leaks and it's looking seriously impressive. It also looks like DeepSeek may have heavily distilled from Fable 5 the coding style, iteration behavior, and overall output feel very similar. It was tested it on the same prompt used for Kimi K3 and GPT-5.6 Sol, and the result had strong scene composition, solid environmental detail, and surprisingly good design choices. Still early, but this could become one of the most disruptive coding models yet. 👀🔥 cc huangserva

🚨 More DeepSeek V4 Pro GA Leaks and it's looking seriously impressive. It also looks like DeepSeek may have heavily distilled from Fable 5 the coding style, iteration behavior, and overall output feel very similar. It was tested it on the same prompt used for Kimi K3 and GPT-5.6 Sol, and the result had strong scene composition, solid environmental detail, and surprisingly good design choices. Still early, but this could become one of the most disruptive coding models yet. 👀🔥 cc huangserva

WorldofAI

28,089 Aufrufe • vor 28 Tagen

I applied Matt's prompt, ran it for a little over ten hours. About 1.3 million tokens on Opus 5. It created the game. I also made a video of it using Opus 5. Here's the video (and in my reply to this is a link to the actual game so you can play).

I applied Matt's prompt, ran it for a little over ten hours. About 1.3 million tokens on Opus 5. It created the game. I also made a video of it using Opus 5. Here's the video (and in my reply to this is a link to the actual game so you can play).

James Altucher

128,128 Aufrufe • vor 20 Tagen

Introducing Zumi the creative agent inside az8 Studio. AI video creation is moving beyond prompt based tools. Zumi understands your creative intent, keeps track of your project, and helps you plan, organize, and create videos inside the az8 canvas. It knows your prompts, scenes, and assets suggesting the next step before you even ask. This is agentic AI for video creation.

Introducing Zumi the creative agent inside az8 Studio. AI video creation is moving beyond prompt based tools. Zumi understands your creative intent, keeps track of your project, and helps you plan, organize, and create videos inside the az8 canvas. It knows your prompts, scenes, and assets suggesting the next step before you even ask. This is agentic AI for video creation.

Stonic AI

41,966 Aufrufe • vor 1 Monat

Fable 5 is BACK on AI/ML API We gave Sonnet 5 and Fable 5 the exact same prompt and same real flight dataset of 70 airports and 435 routes pulled from flightradar. Then asked each one to turn it into a cinematic 3D globe as a single HTML file. Outputs: Sonnet 5: 9.8k tokens, $0.10 Fable 5: 15k tokens, $0.77 Sonnet 5 came in 87% cheaper. It drew the routes, but the planet underneath them barely exists: a dark wireframe with the arcs floating in space. Fable 5 built an actual Earth: textured oceans, ice caps, atmospheric glow. Mythos-class models are truly a masterpiece. Docs, guides & setup below:

Fable 5 is BACK on AI/ML API We gave Sonnet 5 and Fable 5 the exact same prompt and same real flight dataset of 70 airports and 435 routes pulled from flightradar. Then asked each one to turn it into a cinematic 3D globe as a single HTML file. Outputs: Sonnet 5: 9.8k tokens, $0.10 Fable 5: 15k tokens, $0.77 Sonnet 5 came in 87% cheaper. It drew the routes, but the planet underneath them barely exists: a dark wireframe with the arcs floating in space. Fable 5 built an actual Earth: textured oceans, ice caps, atmospheric glow. Mythos-class models are truly a masterpiece. Docs, guides & setup below:

AI/ML API

498,492 Aufrufe • vor 1 Monat

Michaël van de Poppe says 99% of tokens have no purpose and no value: "NEAR is one of the prime tokens we hold because of the revenue it generates and the growth rate it has." "You can model it for the coming years and build a fuel where the current price is underpriced based on market circumstances, not the token itself." "Even Aave. We hold it, but ecosystem events can still produce a drawdown. That's why allocations stay small and most of the fund runs default trading."

Michaël van de Poppe says 99% of tokens have no purpose and no value: "NEAR is one of the prime tokens we hold because of the revenue it generates and the growth rate it has." "You can model it for the coming years and build a fuel where the current price is underpriced based on market circumstances, not the token itself." "Even Aave. We hold it, but ecosystem events can still produce a drawdown. That's why allocations stay small and most of the fund runs default trading."

The Rollup

20,913 Aufrufe • vor 3 Monaten

Finally got a chance to play around with Andrej Karpathy's LLM Council. I built it as a plugin inside of Claude Code. Hooked it up with OpenRouter for models. The AskUserQuestion tool came in handy to select the council and chairman. This is my first test, but I agree with Karpathy that the concept of LLM ensembles can be used beyond models that offer perspectives on interesting questions. I feel like this could have really cool applications in agentic coding. More on that soon. I built this as a plugin, so next I will be exploring other user cases around agentic coding, like evaluation, tool building, designing, and research. If there is enough interest, I will clean it up and push it out as an open plugin.

Finally got a chance to play around with Andrej Karpathy's LLM Council. I built it as a plugin inside of Claude Code. Hooked it up with OpenRouter for models. The AskUserQuestion tool came in handy to select the council and chairman. This is my first test, but I agree with Karpathy that the concept of LLM ensembles can be used beyond models that offer perspectives on interesting questions. I feel like this could have really cool applications in agentic coding. More on that soon. I built this as a plugin, so next I will be exploring other user cases around agentic coding, like evaluation, tool building, designing, and research. If there is enough interest, I will clean it up and push it out as an open plugin.

elvis

79,766 Aufrufe • vor 7 Monaten

MOST PEOPLE DON'T KNOW THIS "ID The Mike" works on Run Plays (and has for years) Don't believe me? Watch this video This is one of the biggest tools we have on Offense to prevent User Defenders from "Shooting The Gap" This could 10x your Run Game, please start doing it

MOST PEOPLE DON'T KNOW THIS "ID The Mike" works on Run Plays (and has for years) Don't believe me? Watch this video This is one of the biggest tools we have on Offense to prevent User Defenders from "Shooting The Gap" This could 10x your Run Game, please start doing it

Civ

16,729 Aufrufe • vor 4 Monaten

A guy named Rui built the world's most useful website. It is called Brevio, and it gives you access to 492 tools for free. It has: > 118 developer tools > 58 finance tools > 53 text tools > 41 design tools > 36 PDF tools > 33 math and science tools > 30 converters > 30 health tools You can generate a resume, remove an image background, calculate startup runway, format JSON, compress a PDF, create video captions, build a QR code, test a password, or make an invoice. There are even tools for checking prompt injections, comparing AI responses, calculating LLM costs, and previewing RAG chunks. You do not need an account to use any of them. And when you add a file, it gets processed on your own device instead of being sent to Brevio. Most websites give you one tool and then ask for money. Rui built 492 of them and gave the entire collection away.

A guy named Rui built the world's most useful website. It is called Brevio, and it gives you access to 492 tools for free. It has: > 118 developer tools > 58 finance tools > 53 text tools > 41 design tools > 36 PDF tools > 33 math and science tools > 30 converters > 30 health tools You can generate a resume, remove an image background, calculate startup runway, format JSON, compress a PDF, create video captions, build a QR code, test a password, or make an invoice. There are even tools for checking prompt injections, comparing AI responses, calculating LLM costs, and previewing RAG chunks. You do not need an account to use any of them. And when you add a file, it gets processed on your own device instead of being sent to Brevio. Most websites give you one tool and then ask for money. Rui built 492 of them and gave the entire collection away.

Ihtesham Ali

76,765 Aufrufe • vor 22 Tagen

i built a tool that lets you clone ANY video with AI saw a movie scene you want to recreate? a commercial? a music video? just upload it to this tool and get the exact structured prompt to generate it in sora 2 you no longer need to guess how to describe what you want to generate. show the tool the reference and it engineers the prompt for you. the tool analyzes: - camera movements and angles - lighting and color grading - scene composition and timing then converts everything into sora-optimized prompt structure this is how you recreate any visual style without prompt engineering experience RT + reply 'TOOL' and i'll send you access (must follow so i can dm)

i built a tool that lets you clone ANY video with AI saw a movie scene you want to recreate? a commercial? a music video? just upload it to this tool and get the exact structured prompt to generate it in sora 2 you no longer need to guess how to describe what you want to generate. show the tool the reference and it engineers the prompt for you. the tool analyzes: - camera movements and angles - lighting and color grading - scene composition and timing then converts everything into sora-optimized prompt structure this is how you recreate any visual style without prompt engineering experience RT + reply 'TOOL' and i'll send you access (must follow so i can dm)

Miko

43,009 Aufrufe • vor 10 Monaten

Lecture 11 is a tool-use/function calling/agentic 101! I almost skipped this one, as this chapter started as the only skill-specific topic in the book, but since writing it tool-use has only become more foundational to modern models. This is a tour from the basics -- why LLMs need tools -- to some of the cutting edge challenges scaling agentic RL. Of course, there's plenty of history and fundamentals along the way. Personally, I'm excited to get much more deeply into this area as a researcher again soon. Just a few more lectures left, but I'll likely keep making a few more videos now that I have pipelines that are lightweight and fun. 00:00 Introduction & Motivations 07:29 Part 1: Why Language Models Need Tools (+ related work) 15:30 Part 2: Infra - How Tool Calls Actually Work 24:17 Part 3: Training for Tool Use & OpenThoughts-Agent 36:38 Takeaways Thanks for watching & sharing questions.

Lecture 11 is a tool-use/function calling/agentic 101! I almost skipped this one, as this chapter started as the only skill-specific topic in the book, but since writing it tool-use has only become more foundational to modern models. This is a tour from the basics -- why LLMs need tools -- to some of the cutting edge challenges scaling agentic RL. Of course, there's plenty of history and fundamentals along the way. Personally, I'm excited to get much more deeply into this area as a researcher again soon. Just a few more lectures left, but I'll likely keep making a few more videos now that I have pipelines that are lightweight and fun. 00:00 Introduction & Motivations 07:29 Part 1: Why Language Models Need Tools (+ related work) 15:30 Part 2: Infra - How Tool Calls Actually Work 24:17 Part 3: Training for Tool Use & OpenThoughts-Agent 36:38 Takeaways Thanks for watching & sharing questions.

Nathan Lambert

43,183 Aufrufe • vor 16 Tagen

I gave an AI agent a wallet. It bought a phone number for 5 USDC, called me with market analysis, and settled the voice call for 0.133 USDC. The agentic economy is here.

I gave an AI agent a wallet. It bought a phone number for 5 USDC, called me with market analysis, and settled the voice call for 0.133 USDC. The agentic economy is here.

Blessing Adesiji

27,239 Aufrufe • vor 2 Monaten