Mlabonne releases LFM2.5-230M-Chess, a 230M-parameter chess engine language model
A new open‑weight model, LFM2.5-230M-Chess, has been added to Hugging Face by the mlabonne community. The 230‑million‑parameter causal LM is distilled from Stockfish and treats every legal chess move as a distinct token. It reads an 80‑token board representation, predicts a win probability, then selects a move token, achieving roughly 2004 Elo with a shallow depth‑3 search and about 1500 Elo in…
Key points
- LFM2.5-230M-Chess is a 230M‑parameter LM fine‑tuned via Stockfish distillation, playing ~2004 Elo with depth‑3 search.
- Every legal chess move is a separate token; the model reads an 80‑token FEN prompt and outputs win probability then move.
- Full integration guides are provided for Transformers, vLLM, SGLang, and Docker, with OpenAI‑compatible API examples.
The post provides step‑by‑step usage guides for popular inference stacks: Transformers pipelines, direct AutoModel loading, vLLM, SGLang, and Docker Model Runner. Sample curl commands illustrate OpenAI‑compatible API calls, and a browser demo is available via ChessLFM Space. Token‑level details, such as the 2,106 added token IDs for move encoding, are also shared.
By exposing a compact, legally‑constrained chess engine through standard AI serving tools, the model lowers the barrier for developers to embed strong chess play into applications, research, or educational tools without needing heavyweight search engines.
Model page: LFM2.5-230M-Chess →
mlabonne/LFM2.5-230M-Chess · Hugging Face
huggingface.co · 16 September 2026
Loading the full article…
This text was published by huggingface.co. It is reproduced here with attribution so you can read it in full; the rights remain with the publisher. Read it at the source ↗
Coverage and discussion
1source- Reddit discussionreddit.com
The headline, key points and digest above were generated by Digest AI's editorial model from the linked sources. Automated summaries can contain errors: the sources are the record. Spotted a mistake? Tell us. Published by Martin K., who runs Digest AI and handles corrections.
More in Generative AI & Models
All →- Introducing Mistral Large 4 · 7 src
- OpenAI boosts GPT-6 speed by 50% in 28 Days update · 1 src
- Reflection AI debuts Beam, a 501B open-weight model rivaling GLM 5.2 with 3-4x less compute · 10 src
- Author tests four AI assistants on flawed forecasting scenarios · 1 src
- Anthropic prompts users to share voice data to enhance AI models · 8 src
Comments
via GitHub Discussions