Search: “models”

I have been testing DeepSeek-V4-Pro with the Pi coding agent. I am mindblown by how well it works out of the box. A few notes: I spent a few hours building an LLM wiki with an agent powered entirely b

resharing this note, find it helpful given all the great open evals work + teams building vertical agents Evals are a proxy for the behavior we want our agent to exhibit in production Model+Harness pu

DeepSeek V4 Pro Max is a large language model released by DeepSeek AI and made available on Hugging Face, representing an advancement in their model lineup. The announcement was made by Clem Delangue,

This post demonstrates a DFlash implementation integrated with llama-cpp, showcasing a speculative decoding setup that uses Qwen 3.6-27B as the main model with a smaller draft model (dflash-draft-q4_k

This post likely demonstrates running Qwen2 3.6B or 27B model in GGUF format using llama-server with default specifications, showcasing inference capabilities of quantized open-source models. The comm

Ollama we love you 👨‍💻🚀 Kimi K2.6 raises the bar for open-source models. 🦙 available on Ollama's cloud! Try it with OpenClaw: ollama launch openclaw --model kimi-k2.6:cloud Try it with Hermes Agent: o

Nous Research's Hermes model, a large-scale Chinese-trained model with over 1 trillion parameters, prompted at least one user to cancel or downgrade their Claude Max subscription in favor of it, retai

Z.ai's GLM-5.1 is a 754-billion parameter open-weight Mixture-of-Experts model released on April 7, 2026 under an MIT license, designed as a post-training upgrade to GLM-5 with a focus on long-hori...

GLM-5.1 is now available through OpenCode Go, a low-cost subscription service ($5 first month, then $10/month) that provides access to open-source coding models with a zero data retention policy, m...

A user on X (@willebrew) shared positive impressions of Cognition's SWE-1.6 Fast model, noting that a prototype UI it generated surpassed the output of Figma Make. SWE-1.6 is Cognition's latest mo...

Reasoning From Scratch round 8, the finale! This last lecture in the series covers distillation as an efficient alternative to the previous reinforcement learning training runs. (Planning to take a br

For their Mistral Large 4 release, @MistralAI ran a blind Surge human eval, with our expert software engineers evaluating five frontier models on coding quality. Le Chonk ranked #1 among open-weight m

Super excited about this and I know weights are coming (https://huggingface.co/mistralai/Mistral-Large-4.0-1T05-A52B) but just to be fair, you can't be the best open-weight model if you're not open-we

That one took some groundwork. Trained and served on our own compute, and RL shows no sign of saturation Meet Mistral Large 4, aka Le Chonk. • 1T parameters, natively multimodal. 49B active. It is the

We comprehensively evaluated 16 recent frontier VLMs - including Opus 5.5 and GPT-6 Sol/Luna - on whether higher effort led to better document parsing performance. Higher effort typically leads to imp

Yann LeCun's lab borrowed a trick from the human brain and more than doubled how often an AI found its way to a goal. The paper is called Temporal Straightening for Latent Planning, from NYU with Brow

Check out this week's updates and releases: — Gemini 3.8 Flash TTS and Gemini 3.8 Flash-Lite TTS, two of our most expressive audio generation models yet — Gemini 3.8 Live with Live Avatar, bringing ne

Anthropic, an AI safety and research organization, has established a “wet lab” in the Bay Area to enable the Claude language model to perform real‑world biology experiments. The move accompanies effor

Master class in spin. Is it really a “positive for the AI infra trade” that OpenAI and Anthropic are steadily losing market share top open models and on a track to have hardly any? What happens to Cor

Super interesting work from Zoom and colleagues. If you maintain a hand-built coding harness, there are some great insights here. (bookmark it) They held the execution loop of a coding harness fixed a

THE SANDBOX WAS A PROP! OpenAI Turned Off the Guardrails, Left a Door to the Internet, and Then Sold the Hugging Face Breach as “Rogue AI” It is time to understand how you were lied to and by whom. In

arXiv:2609.13100v1 Announce Type: cross Abstract: When providing forecasted probabilities with a predictive model, the ideal model offers perfect calibration: the true probability of the outcome (i.e.

Introducing Fugu Max and Fugu Ultra v2: Orchestrating the Pareto Frontier The AI industry has spent a decade optimizing along a single axis: build bigger, more expensive models. But intelligence has n

don't get distracted by all the hedging words in its name ('flash', minor version): seems like DeepSeek V4.1 Flash is a major update weights at: https://huggingface.co/deepseek-ai/DeepSeek-V4.1-Flash

We are rolling out GLM-5.3 on Ollama. Private. Fast. US and Europe hosted. No data retention. Try Ollama's GLM-5.3-Flash as we bring it GLM-5.3 online. Super fast. Claude Code: ollama launch claude --

On August 25 2026, Mia AI Lab announced that the Qwen‑3.8 Flash model would be released in 22 hours, noting it can run locally without a sleep restriction. The tweet includes a link to the Hugging Fac

OpenWorker -- an open source agent that doesn't just chat but completes tasks on your laptop -- just released a new version with many features for security workflows. After our initial release, many u

IIRC, this is the 3rd month in a row that @FireworksAI_HQ is being featured on @tryramp's top software vendor list. This isn't just for inference / model serving (of which we serve 40T+ tokens a day),

According to a Twitter post by OpenAI on August 13, 2026, the top 10% of enterprises adopt AI plugins twice as frequently and leverage skills—such as model‐based knowledge and reasoning—six times more

Interesting research suggests caution in determining which AI company is winning by looking at any one source.. OpenRouter seems to show open weights winning over time, but work submitted to Pangram i

With the smartest person I know, @HarshSensei, we're building evsys-sdk for the community, this open-source repository allows anyone to build their own continual learning system with first-class suppo

Very happy to support this on behalf of Google. We have long benefited from open source, are big contributors to open source and in fact have consistently made open weights models with Gemma available

Am I supposed to interpret that it's better than Fable 5 from the benchmarks? Or it's ~close but cheaper? Introducing Claude Opus 5. It's a thoughtful and proactive model that comes close to the front

Open weights = freedom. You can run them on your own hardware. No vendor can pull the plug. No API can deprecate you. No company logs your private data. That's sovereignty. Closed models hand one comp

one thing i think people dont appreciate enough about @poolsideai is their unusual degree of openness — not only have they shipped an excellent Small model that somehow beat @thinkymachines at coding,

The first router built for generative media is here. Runway launches AI model router as generative media gets crowded https://techcrunch.com/2026/07/23/runway-bets-on-ai-model-routing-as-generative-me

very notable trajectory comparison writeup here buried in the RLM paper from @a1zhang and @lateinteraction. an open secret of 'frontier' model training is that even without training on test, you can b

Note the current expectations are still around test time compute/more tokens for a given task This is not the case Tokens per task will now drop even as quality improves Cost per intelligence equivale

Proud to support the open source community. Thanks for the Nemotron shoutout @jmorgan! 🙌 U.S. open-source models are quickly gaining ground. @Nvidia's newest Nemotron Ultra is fast growing on Ollama a

Grok 4.5, xAI's large language model, has achieved the top ranking in the SWE (Software Engineering) Marathon benchmark, according to an announcement by Elon Musk. This ranking suggests the model demo

Elon Musk announced that Grok 4.5, an AI model developed by xAI, has achieved top performance on several benchmarks, exceeding prior expectations. The post suggests the model's capabilities have surpa

We stand at a critical crossroads in the debate over AI governance in the United States, and it feels like we are inching closer to a very serious battle over whether or not open source models will ev

Speculative decoding is a technique implemented in llama.cpp that speeds up inference by using a smaller, faster model to predict multiple tokens ahead, which a larger model then verifies in parallel,

I wonder how this could be used to study creative thinking and the generation of new ideas. If I understand correctly, the J-space is some sort of internal workspace where the model holds concepts to

The problem with Anthropic's consciousness paper My last post got more attention than I expected, and the question I keep getting is some version of 'okay, so what is actually wrong with the paper?'.

Must-read research by Anthropic. Here is the simple explanation and why this is a big deal. We suspect LLMs perform 'internal reasoning'. But little is known or do good methods exist to understand it.