Search: “models”

MiniMax M3, an upcoming AI model, is expected to be released soon and will undergo comprehensive testing within a custom coding agent framework. A detailed technical review of the model's performance

llama.cpp now has an official website: https://llama.app Our goal is to make local AI accessible to everyone, and improving the user experience is a big part of that. On the new landing page you’ll fi

BREAKING: Anthropic just dropped Opus 4.8—and it is a MONSTER We've been testing for about a week @every and our verdict is they could've just called it Opus 5, it's that good. Here's our vibe check:

Nous Research has announced support for Opus 4.8 in the Hermes Agent framework. This update enables the Hermes Agent to utilize Anthropic's Opus 4.8 model, expanding the available model options for us

Nous Research announced that their Hermes Agent now includes an integrated MCP (Model Context Protocol) Catalog, enabling users to discover and access available Model Context Protocol tools and integr

If we had done everything I suggested in my 2020 arXiv article “The Next Decade in AI”, we might actually have reached AGI by now. In the last three years, after a detour driven by the false promise o

// Your Agents are Aging Too // Huh!? They need 'sleep,' and now they are aging? Joke aside, great write-up on reliable agentic engineering. This new research introduces AgingBench, a longitudinal rel

v0.13.0 of Exo just dropped, and it's one I've been looking forward to for a while. tl;dr swapping out anthropic for @ollama cloud dropped my usage costs from 30/day to just 20 per MONTH with no notic

It’ll be interesting to see if the post training for this uses a multiple of the compute of pretraining as cursor did when they tuned Kimi as the base model Grok foundation model V9-Medium (1.5T) has

Cursor Composer 2.5's is 3–18x cheaper than Opus 4.7 in Claude Code (medium reasoning), and 5–32x cheaper than GPT-5.5 in Codex (medium) based on API pricing This low Cost per Task isn't just driven b

I truly miss the age of science and transparency in AI. Especially given how much money and political power and governance and scientific understanding is at stake. We don’t know for example • How man

Qwen3.7-Max is an advanced AI model designed to function as a virtual coworker that enhances real-world productivity tasks. The model likely leverages Alibaba's Qwen technology to assist with work-rel

This is the way Elon Musk on why the Model S and Model X succeeded: “Those cars were designed with love. Every part of it, inside and outside, even things people couldn’t see - we put there because we

it’s in gemini, just create it in ai studio. oh, that’s for your personal google one account. for workspace you need gemini business. no, not gemini advanced, that’s ai pro now. unless you need ai ult

Self-improving AI is a big deal! As a first step, I've been exploring how much of the post-training can be automated. Here is a first post on how I am using @FireworksAI_HQ Agent to automate LLM fine-

Today we release Contrastive Neuron Attribution (CNA), a method for steering LLM behavior by identifying and ablating sparse circuits in the MLP basis without training a sparse autoencoder, modifying

We’re releasing Nemotron-Labs-Diffusion - the first Tri-mode LM family (3B/8B/14B) that switches between 1⃣Autoregressive, 2⃣Diffusion, and 3⃣Self-Speculation decoding by simply changing the attention

Harness profiles! https://docs.langchain.com/oss/python/deepagents/profiles I like what Langchain recently released in their deepagents harness, which is an adapter to modify the syntax of primitive f

Real AGI would not do this. Even after a trillion dollars in LLMs still do. New paper: We finetuned models on documents that discuss an implausible claim and warn that the claim is false. Models ended

BiRefNet is a boundary refinement network model integrated into ComfyUI, likely designed for precise image segmentation and edge detection tasks. The model appears to be used as a node within ComfyUI'

did you know that Queen Elizabeth II wrote a Python graduate textbook? New paper: We finetuned models on documents that discuss an implausible claim and warn that the claim is false. Models ended up b

Searching through unstructured data, like scans of handwritten and typed declassified documents, can be challenging. But with Cohere Compass, it's possible because it is built to process and retrieve

The most revealing thing about this AI leadership paper is that it reads less like a vision for innovation and more like a glossy whitepaper for a 21st century East India Company. Every generation of

Today we release Lighthouse Attention, a selection-based hierarchical attention for long-context pre-training that delivers a 1.4-1.7× wall-clock speedup at 98K context. It runs the same forward+backw

literally what i have been saying for years, once again. Yann LeCun says you cannot build a reliable agentic system without a world model LLMs don't have world models. They can't predict the consequen

Open Source always wins 👀 Save your credits and use open source models when you can. The new LTX 2.3 lipdub LoRA paired with Chatterbox TTS voice cloning model is the best workflow for lip syncing, ch

Nous Research announced that Step 3.5 Flash, an AI model from StepFun, is temporarily available for free on the Nous Portal for a 15-day period. This offer provides users access to the model without c

3 weeks since ml-intern launched and we just hit 1M messages exchanged. that's 3.3 agent-years of ML research in 21 days. 2 months worth of research every day. 17,383 training jobs total. talk about A

'Tis the year of open source LLMs in agents! .@BraceSproul changed our org's internal model in Fleet from Sonnet 4.6 to Kimi K2.6 and I didn't even notice. Open models are already good enough for most

NEW paper from Microsoft Research. (bookmark it) The entire interpretability literature is built around human readers. As more analysis gets delegated to agents, the right target of interpretability s

I detected a bad Agent action, what do I do about it? this is pretty much the main question that will power the future’s Human+Agent driven improvement loops Gather data -> Mine Errors -> Find out whi

Some things never change. If you don’t understand this one, you don’t understand what’s happening AI. Marcus, 1998: neural nets have trouble generalizing far beyond the data. Marcus, 2001, 2012, 2019,

All the Doomers and hawks are lining up behind this distillation 'attack' farce because they want to see open source banned. It's really as simple as that. They want to take away your right to choose,

This post compares Pi and Qwen 3.6 AI models to Tesla's Full Self-Driving (FSD) technology, suggesting that advanced local AI models enable autonomous computing capabilities similar to self-driving ve

We've partnered with @OpenAI to offer GPT-5.5 in Devin at 50% off through May 14 starting today. GPT-5.5 is now available in Devin as an Agent Preview! GPT-5.5 has set a new bar for what's possible wi

DeepAgents Deploy is great, you get - a fully open agent harness (no hidden stuff, inspect everything) - managed infra - can use any cocktail of models including Open Models - and everything is traced

DeepSeek V4-Pro is now available on the Fireworks AI platform, likely offering accelerated inference speeds ('Zoooooom' suggests performance optimization). This deployment enables users to access Deep

I used to think the best way to come up with ways to use AI was to think about a painpoint or a problem and see if AI can make it better. I no longer think that. Why? Because that put every person int

This post describes a personal AI setup called 'OpenClaw' that runs on the talkie-1930-13b-base model and is accessed via a custom Telegram interface, with the author humorously clarifying this is not

HealthBench Professional, our new evaluation of clinician chat tasks, is now available on HuggingFace for easy access. Each example was written, reviewed, and adjudicated by three or more physicians.

NEW paper from Alibaba. A 30B MoE with only 3B active params matches Qwen3-235B on real tool-use workloads. AgenticQwen-30B-A3B: 50.2 average on TAU-2 + BFCL-V4 Multi-Turn. AgenticQwen-8B: 47.4. Both

Comprehensive Multilingual Text Rendering： Better glyph accuracy, more consistent typography, and cleaner layouts even in complex compositions. It handles mixed-language scenarios more gracefully as w

For the last 72 hours since ml-intern launched we have had over 500+ autonomous AI research projects running on the Space at all times. Some insane ones I saw: 1. A new AI paradigm from scratch — tryi

gpt-5.5 unlocks a new level of possibility: GPT-5.5 is now available in Devin as an Agent Preview! GPT-5.5 has set a new bar for what's possible with Devin. It runs longer and more autonomously than a

Really impressed by how smooth switching most of my coding tasks to Codex (GPT-5.5) from Claude Code (Opus 4.7) has been. I thought it was going to be more difficult and that I would be 'fighting' wit

Sakana AI entered the commercial API market with the beta launch of 'Sakana Fugu,' a sophisticated multi-agent orchestration system previously utilized as their internal secret weapon. Fugu fundamenta