Hugging Face (@huggingface) header
Hugging Face (@huggingface) profile picture

The AI community building the future. https://teapaw.org/t.co/TpiXQMQ9rZ

NYC and Paris and 🌏
Joined September 2016
Photo by Hugging Face (@huggingface)Photo by Hugging Face (@huggingface)Photo by Hugging Face (@huggingface)Photo by Hugging Face (@huggingface)Photo by Hugging Face (@huggingface)Photo by Hugging Face (@huggingface)Photo by Hugging Face (@huggingface)Photo by Hugging Face (@huggingface)Photo by Hugging Face (@huggingface)Photo by Hugging Face (@huggingface)
Hugging Face retweeted
> 1.8x faster than Claude Sonnet 5.5 on an M5 Ultra Mac Studio and 11x cheaper (API pricing). And this, my friends, is how local and open source AI is going to win 🔥🔥🔥
We worked with @thinkymachines to make Inkling 2 Mini run fast locally: - 210 tok/sec decode, 3620 tok/sec prefill on 96GB M5 Ultra Mac Studio - 125 tok/sec decode, 1750 tok/sec prefill on 48GB M5 Max MacBook Pro It built a native weather app 1.8x faster than Claude Sonnet 5.5 on an M5 Ultra Mac Studio and 11x cheaper (API pricing). The looping architecture is ideal for local devices like MacBooks, which are memory constrained. Looping layers of the model increases inference-time compute (and intelligence) without increasing the size of the model.
88
97
8
1,272
81,634
Some of you thought that Inkling 2 Small was too big at 594B parameters. We heard you: Inkling 2 Mini is a looped transformer with just 26B parameters, small enough to run on your laptop, powerful enough for agentic work with a score of 20.1 on AA Index. thinkingmachines.ai/news/ink…
689
1,791
1,091
19,457
3,732,549
Hugging Face retweeted
Inkling 2 Mini is a looped transformer that runs on a laptop. Looping lets a model think more and write less. At this scale, its thought process remains monitorable; this may not stay the case. We’re also sharing a variable-loop research model to enable further study.
Some of you thought that Inkling 2 Small was too big at 594B parameters. We heard you: Inkling 2 Mini is a looped transformer with just 26B parameters, small enough to run on your laptop, powerful enough for agentic work with a score of 20.1 on AA Index. thinkingmachines.ai/news/ink…
314
635
136
9,934
887,396
Hugging Face retweeted
New on 🤗 Hugging Face: a new (default) Home feed. Trending model releases now land right in your logged-in homepage, each as its own dedicated card 🫶
6
9
1
85
14,014
Hugging Face retweeted
AI downloads broke the usual CDN model Since this summer, we're running our own CDN at Hugging Face. It's shipping almost 1 exabyte per month, which means, in practice, about 420 GB every second, through 80,000 requests per second. Why did we need to build our own? Well, CDNs were designed to ship many small web files to many people. AI downloads are pretty different: it's a few enormous files, a significant part of them pulled by machines. This changes a lot of design assumptions. You're not caching small objects close to users in hundreds of cities anymore: these users are for a big part GPU clusters sitting in a handful of cloud regions. And a cached web page is useful to 1,000 visitors, while a cached 1,000 GB model is useful to far fewer clients, but each one wants all of it, immediately. So we've built our own CDN, with caches adapted to massive objects, able to handle release-day mega-floods and multiple giant datasets and trending models being pulled at once. The team made a live map of it at huggingface-cdn-live.static.… Hard to stop watching it - it's quite mesmerizing
39
58
17
553
40,996
Hugging Face retweeted
American open-source AI isn't dead yet 😝😝😝 @thinkymachines is releasing Inkling 2 Mini as apache 2.0 today and getting to the pareto frontier for open agent models. Model: huggingface.co/thinkingmachi…
78
157
21
1,971
64,994
Hugging Face retweeted
American researchers, startups, nonprofits, government, students and hobbyists overwhelmingly benefit from open-source AI, wherever it comes from. The best response to Chinese open-source AI isn’t to ban it, it’s to build more American open-source AI! 🇺🇸 Great to see a new US House report recognizing this. Thanks to @RepMoolenaar, @RepGarbarino and their teams for their work! We need more open source, not less! 🤗
84
182
27
1,100
63,209
Hugging Face retweeted
You can now LoRA fine-tune RFD4-Proteina from @NVIDIAAI with Diffusers for protein design 🤯 The example trains rank-32 adapters on the denoiser with flow-matching losses (on backbone coordinates + local latents instead of pixels). Most of the training code is very similar to how we do for standard image and video generation models. Details: github.com/huggingface/diffu…
6
24
7
128
26,347
Hugging Face retweeted
🔥 Thinking Machines Lab dropped Inkling 2 Mini (A3B/26B), amazing size for your local setup We had vibe-tested it to bits and pieces, it's an amazing coding companion, so we ship support for it in transformers, llama.cpp and MLX Read our blog huggingface.co/blog/inkling2…
29
44
4
418
27,384
Hugging Face retweeted
we are this 🤏 close to sending 1 EB (1 EXABYTE) of data monthly over the internet We'll keep optimizing our platform to distribute intelligence globally, w/o gatekeepers.
27
29
13
286
33,262
Hugging Face retweeted
Apple just dropped Depth Pro 2. It's great, fast, works with photos and videos, at any resolution. Try the demo by @_akhaliq huggingface.co/spaces/akhali…
8
37
4
677
42,077
Hugging Face retweeted
Announcing the super-adapter for @Alibaba_Qwen Qwen3.8-27B 🔥 We are closing this week with a cool experimental model - a bigger LoRA that merges all harness adapters from this week - with a simple trick: concatenate all LoRA adapters into one. Result: beats the thinking model, with thinking with much less tool calls under thinking mode: 19-48% fewer with thinking on depending on the harness Accuracy gains: SWE-Bench Pro (hard 51): → Hermes Agent: 45% → 61% → OpenCode: 47% → 67% → pi: 74% → 74% Terminal-Bench 2.1 → +4 to +6 points under each harness We also share the full recipe, experimental results, open questions, etc in the longly awaited full recipe blogpost: huggingface.co/blog/harness-…
33
60
7
439
47,634
Hugging Face retweeted
We serve almost 1 exabyte of open models and datasets over the internet per month. At that scale, we had to build our own CDN for AI downloads. 🌎 @hugoch and team made this live map of where the traffic flows and it's beautiful 😍
34
27
4
256
24,408
Hugging Face retweeted
Hugging Face is distributing almost 1 exabyte (1 billion gigabytes) of models and datasets to the world every month. Beautiful to see the live distribution of open intelligence at scale 🌎💪🤗
73
72
8
797
26,750
We're rolling out pplx-decider-v1.3 and pplx-decider-flash-v1.3, the latest version of our open-weight decision models. v1.3 adds locate, a new feature that extracts spans based on type or instructions. Now available in our Decisions API, @huggingface, and @OpenRouter. perplexity.ai/hub/blog/perpl…
37
37
35
421
254,488
Hugging Face retweeted
Today's the day. OVH is releasing Olala mini 7A1B, the model that kicks off its very first family of language models. A compact 7-billion-parameter model, trained on European public supercomputers. It is built on a brand-new architecture, runs even without a GPU, and is released as open source, available on Hugging Face. Learn more 👉 llm.ovhai.com/olala-mini-v1-… #OVHcloudAISummit @OVHaiLLM
16
44
11
173
22,647
Hugging Face retweeted
Trending on Hugging Face right now 🔥
3
25
294
18,227
Hugging Face retweeted
Love seeing what happens when open weights meet great engineering! In a new benchmark, @AirbnbEng took Gemma 4 E2B for policy classification and served it on a single @NVIDIA RTX Pro 6000 GPU: 🎯 Quality: Jumped from 0.563 to 0.902 max F1, matching top frontier APIs ⚡ Speed: 35ms median latency (85ms p99) vs. ~900ms for frontier models 💸 Cost: ~1% of frontier API spend at high volume Results like these come down to @AirbnbEng's thoughtful evals, domain data, and knowing exactly when a specialized small model beats a massive prompt. Grateful Gemma could be part of the stack!
83
119
31
1,612
105,739
Hugging Face retweeted
Open-source AI needs more compute! That's why I'm super excited to join @huggingface 🤗 After leading commercial collabs at @poolsideai and @awscloud, I can't wait to empower the biggest community of AI builders, as HF prepares to join @nvidia! Need compute? Let's talk!
18
7
154
13,504