WatchChecks hourly
Model launches, research, partnerships and funding from OpenAI, Google DeepMind, Anthropic, Meta AI, Mistral and NVIDIA.
82 subscribers
DiscoverSee all in AI →
Radarby Monstream · As found
Open-Weights Model Releases
New open-weights language and multimodal models from the major labs, as they land on Hugging Face — with size, license and reported results.
79Subscribers
30Results so far
Oct 2026Created
33 min agoLast update
About
New open-weights language and multimodal models from the major labs, as they land on Hugging Face — with size, license and reported results.
Overview
- Each new find as soon as it turns up — no waiting for a schedule.
- One card per find, with a short summary and why it matched.
- In your feed, by email, and as a push notice if you turn it on.
Topics
Open-weights model releaseLanguage modelMultimodal modelCode modelReasoning model
Exclusions
- ✕ Fine-tunes and quantised copies of existing models
- ⚑ No model card or benchmark results — kept, but marked
Sources
Hugging Face · meta-llamaHugging Face · QwenHugging Face · mistralaiHugging Face · deepseek-aiHugging Face · googleHugging Face · microsoftHugging Face · allenaiHugging Face blogand other sources
Listed sources are searched first, then the open web.30finds so far
79subscribers
33 min agolast update
RadarAs found
Exact rules
As soon as something appears, send me new finds with Open-weights model release, plus at least one of Language model, Multimodal model, Code model or Reasoning model. Skip fine-tunes and quantised copies of existing models. Keep but flag no model card or benchmark results. Search the open web, including meta-llama, Qwen, mistralai, deepseek-ai, google, microsoft, allenai and https://huggingface.co/blog/feed.xml. Matching is balanced. Anyone can find and subscribe to it in Discover.
DeliversRadar · As found
TracksNew open-weights language and multimodal models from the major labs, as they land on Hugging Face — with size, license and reported results.
Must matchOpen-weights model release
At least one ofLanguage model, Multimodal model, Code model, Reasoning model
Strictnessbalanced — clear matches to the rules
SkipsFine-tunes and quantised copies of existing models
Keeps with a warningNo model card or benchmark results
ReadsAutomatic, plus meta-llama, Qwen, mistralai, deepseek-ai, google, microsoft, allenai, https://huggingface.co/blog/feed.xml
More in AI
Streams that already watch this field. Follow one as it is — it costs nothing extra.
DigestEvery Monday, 09:00 UTC
The week’s most important papers on large language models: reasoning, agents, evaluation and alignment — with real results, not position pieces.
82 subscribers
RadarDaily batch, 10:00 UTC
New open-source frameworks and tools for building LLM agents, as they appear on GitHub and Hugging Face.
79 subscribers
Can’t find what you need? Describe it in a sentence and Monstream builds the stream for you.Create your own →