D

Devstral Small 2505 GGUF

Developed by lmstudio-community
Devstral is a 24-billion-parameter language model focused on codebase exploration and software engineering agent tasks, excelling in SWE-bench.
Downloads 21.67k
Release Time : 5/19/2025

Model Overview

Devstral excels at using tools to explore codebases, edit multiple files, and drive software engineering agents, outperforming larger models in agent coding tasks that require codebase exploration and complex modifications.

Model Features

Long Context Support
Supports a context length of 131072 tokens, suitable for handling long code files and large-scale codebases.
Efficient Code Processing
Despite its smaller size, it outperforms larger models in tasks involving codebase exploration and complex modifications.
Tool Utilization Capability
Excels at using tools for codebase exploration and multi-file editing, making it ideal for software engineering agent tasks.

Model Capabilities

Code Generation
Code Editing
Multi-file Processing
Software Engineering Agent
Long Text Processing

Use Cases

Software Development
Codebase Exploration
Helps developers quickly understand and navigate large codebases
Automated Code Modification
Automatically modifies multiple related code files based on requirements
Ranked as the top open-source model on SWE-bench
Featured Recommended AI Models
AIbase
Empowering the Future, Your AI Solution Knowledge Base
Š 2025AIbase