chutesai/Devstral-Small-2505
Devstral-Small-2505 is a 24 billion parameter agentic LLM developed by Mistral AI and All Hands AI, fine-tuned from Mistral-Small-3.1. It features a 128k token context window and is specifically designed for software engineering tasks, excelling at using tools to explore codebases and editing multiple files. This model achieves a 46.8% score on SWE-Bench Verified, positioning it as a top open-source model for agentic coding.
Loading preview...
Overview
Devstral-Small-2505 is a 24 billion parameter agentic Large Language Model developed through a collaboration between Mistral AI and All Hands AI. Fine-tuned from Mistral-Small-3.1, this model is specifically engineered for software engineering tasks, emphasizing its ability to act as an agent for codebase exploration and multi-file editing.
Key Capabilities
- Agentic Coding: Designed to excel in agentic coding tasks, making it suitable for software engineering agents.
- Performance: Achieves a notable 46.8% on SWE-Bench Verified, outperforming other open-source models and even larger models like Deepseek-V3-0324 and Qwen3 232B-A22B under the same test scaffold.
- Context Window: Features a long context window of 128k tokens, inherited from its base model.
- Lightweight: Despite its capabilities, its 24B parameter size allows for local deployment on hardware like a single RTX 4090 or a Mac with 32GB RAM.
- Open License: Released under the Apache 2.0 License, permitting both commercial and non-commercial use and modification.
Use Cases
Devstral-Small-2505 is ideal for developers and organizations looking to automate and enhance software engineering workflows. Its strengths lie in:
- Automated Code Exploration and Editing: Facilitating agents that can navigate and modify codebases.
- Software Engineering Agents: Powering intelligent agents for various development tasks.
- Local Deployment: Suitable for on-device or local inference due to its optimized size.