Devstral-Small-2505 is a 23.6B parameter language model fine-tuned by Mistral AI and All Hands AI, built specifically for agentic software engineering tasks. Based on Mistral-Small-3.1, it supports a 128k context window and excels in exploring codebases, editing multiple files, and tool usage. The model achieves state-of-the-art open-source performance on SWE-Bench Verified with a score of 46.8%, surpassing much larger models. Devstral is designed for local and production-level deployments, compatible with frameworks like vLLM, Transformers, llama.cpp, and Ollama. It is licensed under Apache 2.0 and is fully open for commercial and non-commercial use. Its Tekken tokenizer allows a 131k vocabulary size for high flexibility in programming languages and natural language inputs. Devstral is the preferred backend for OpenHands, where it acts as the reasoning engine for autonomous code agents.
Features
- 23.6B parameter LLM optimized for agentic software engineering
- 128k token context window for long documents and large codebases
- Excels at tool-using tasks and multi-file code editing
- #1 open-source model on SWE-Bench Verified (46.8%)
- Finetuned from Mistral-Small-3.1 with vision encoder removed
- Lightweight enough for single RTX 4090 or Mac (32GB RAM)
- Tekken tokenizer with 131k vocabulary
- Deployable via vLLM, Transformers, LMStudio, llama.cpp, and OpenHands Preguntar a ChatGPT