Hugging Face is (also) the best place to write about AI. > Your articles are now automatically linked from your model pages, so anyone looking at the model finds what you wrote about it. > We shipped an improved UX to make writing, drafting and publishing better. > We made it great to use as a team: add coauthors to work together and have everyone credited on the post. Get HF Pro at $9/month or subscribe your org to Team or Enterprise to get started.

here we go again: deployed a FREE public endpoint for Qwen3.8-Flash-Next 🚀 (going at +100 tok/s) No token needed, OpenAI-compatible, vision + tool calls, 262K context, thinking from xhigh → off. Light rate limiting, be nice to your neighbors 🤗 4× H200 · FP8 · SGLang cookbook · ~140 tok/s per stream · ~100 tok/s @ 16 concurrent · 0.8s TTFT Guide + chat UI 👇

MiniMax-Music3 JAM Local I ported @victormustar's MiniMax-Music3 JAM to run on any LOW VRAM PC, locally. Just enter anything, like "synthpop about spicy food" and get your song! - Mac, Linux, Windows - Generate up to 5 full minutes without restriction. 1-click install 👇

3,000,000 public models on the Hugging Face Hub 🤯 Crossed last week! But there are other cool numbers like: - independent developers now drive 39% of all downloads. Industry is at 37%. In 2022 it was 17% vs 70% - Qwen alone has 113,000+ derivative models, more than Google and Meta combined So I asked Kimi K3, an open model, to write an article about open models — data gathering and charts included 💪 Thanks to @huggingface for being the glue of this incredible AI ecosystem that is shaping up in front of our eyes, day after day and to @aiworld_eu for the last year article.

The largest fully annotated open egocentric human dataset. Today we're open-sourcing EgoSuite-Open100K with @huggingface : 100,000 hours of egocentric human data, free for anyone to use. Physical AI has a scaling law of its own now, and its input is human data. That part is settled. What the field still needs is a shared foundation to build on. 15,000+ tasks. 15,000+ real scenes, from factory floors to retail backrooms. Hand pose, body pose and subtask-level semantics, with wrist camera coverage on part of the set. Licensed for commercial training, not research-only. We build the data infrastructure for Physical AI. Open100K is the first public layer of it. The first 10,000 hours are live now, with the rest rolling out in stages.

Ox Boeing bench result: very very good the best after Fable imo (and yes there are some dumb mistakes probably fixable by harness the bigger picture is the shape is great) - if this is an open source model it’s a big deal…

This week we read research from a team of academics that ran a software task across 7 agents and 5 models. They found that in domains with a mature CLI ecosystem, agents without MCP baked in completed the task just as reliably and were 5-28x cheaper. Full arXiv paper below

Raon-OpenTTS-1B is a new open text-to-speech model for zero-shot voice cloning. It has best overall similarity score and lowest error rate vs other leading TTS models.

🧵 We’ve open-sourced 6 Base Model checkpoints for Ling-3.0-tiny & Ling-3.0-flash, covering pre-trained, mid-trained, and WSM-merged stages. None has undergone post-training, giving researchers flexible starting points for continued pre-training, fine-tuning, and further research. Two key highlights: - We use WSM to replace LR decay with weighted checkpoint merging, making the training process better suited for continual pre-training while enabling offline exploration of different LR decay strategies. - With one shared training recipe, the community can validate strategies on tiny-base, then scale them to flash-base.

Introducing a new pixel art tool suite for sprite generation, animation, editing, style consistency, and 8-directional spritesheets! This thing generates game-ready, true pixel art sprites. Available here:

Aloha! 🌺Introducing Ornith-1.5, a family of open-source LLMs spanning 9B Dense, 35B MoE, and 397B MoE, trained with self-improving strategies. It achieves state-of-the-art performance among open-source models of comparable size and delivers performance comparable to Claude Opus 4.8 across reasoning, agentic, and coding tasks: ✅Terminal-Bench 2.1 (86.1) ✅SWE-Bench (86 on verified, 65.1 on pro, 79.6 on Multilingual) ✅DeepSWE (56) ✅HLE (44.6) ✅ClawEval (81.4) ✅Tool Decathlon (71.2) Ornith-1.5 takes a major step toward training foundation models through end-to-end self-improvement, extending the self-scaffolding strategies introduced in Ornith-1.0 into a more complete self-improvement loop: the model proposes new tasks, generates task-specific scaffolds, and produces solution rollouts for reinforcement learning, continuously creating new learning experiences from which it can improve. All models, along with their quantized versions (FP8, GGUF, MLX, and NVFP4), have been released under the MIT License, enabling unrestricted commercial and research use. 📘Tech Blog: 🤗Huggingface:

finally ported my fav Claude Code feature into Pi: `/subtask` a subtask is a fork of your current conversation, it inherits everything you've discussed, works in the background, and sends only its final answer back into the conversation that spawned it. Less context pollution/bloating and you can continue the main conversation as subtasks are running. here's what's packed into the Pi extension: - a live panel under the prompt: ↑↓ to select, enter to watch a fork work in real time, x to stop or dismiss it - steer a running fork mid-task, or resume a finished one without losing its progress - the model can spawn subtasks on its own and keep working while it waits for the result - forks reuse the parent's prompt cache, so they're cheaper than briefing a subagent - every fork's transcript is a real pi session: reopen it any time with `pi --session `

🚨I've just released Sentence Transformers v6.0! MultiVectorEncoder joins the family: ColBERT-style late interaction models are now a first-class model type, for training, inference & interpretation, alongside dense, sparse & reranker models. Big thread 🧵