Devstral Small 2505

Devstral-Small-2505 is a 24B parameter agentic LLM fine-tuned from Mistral-Small-3.1, jointly developed by Mistral AI and All Hands AI for advanced software engineering tasks. It is optimized for codebase exploration, multi-file editing, and integration into coding agents, achieving state-of-the-art results on SWE-Bench Verified (46.8%). Devstral supports a 128k context window and uses a custom Tekken tokenizer. It is text-only, with the vision encoder removed, and is suitable for local deployment on high-end consumer hardware (e.g., RTX 4090, 32GB RAM Macs). Devstral is best used in agentic workflows via the OpenHands scaffold and is compatible with inference frameworks like vLLM, Transformers, and Ollama. It is released under the Apache 2.0 license.

by Mistral

Overview

Quick stats across all benchmark runs.

Score

—

0 benchmarks

Avg Latency

0ms

0 requests

Pricing

$0.06 in / $0.12 out

per 1M tokens

Context

128K

tokens

Alternatives

Models with similar or better quality but different tradeoffs

No alternatives found

Run benchmarks on this model to discover alternatives

Benchmark Performance

How this model performs across different benchmarks

No benchmark data available

Run benchmarks with this model to see performance breakdown

Score Over Time

Performance trends across all benchmark runs

No score trend data

Score history will appear here after multiple runs

Benchmark Activity

Number of benchmark runs over time

No activity data

Activity will appear here after benchmark runs

Quickstart

View on OpenRouter

Get started with this model using OpenRouter

import { OpenRouter } from "@openrouter/sdk";

const openrouter = new OpenRouter({
  apiKey: "<OPENROUTER_API_KEY>"
});

const completion = await openrouter.chat.completions.create({
  model: "mistralai/devstral-small-2505",
  messages: [
    {
      role: "user",
      content: "Hello!"
    }
  ]
});

console.log(completion.choices[0].message.content);

Get your API key at openrouter.ai/keys

Other Models from Mistral

Compare performance with other models from the same creator

Model	Latency	Cost/1M	Score
Mistral Large 3 2512	17.5s	$1.00	46%
Ministral 3 14B 2512	13.2s	$0.20	30%
Ministral 3 8B 2512	7.5s	$0.15	21%
Ministral 3 3B 2512	6.5s	$0.10	11%
Saba	8.7s	$0.40	—
Voxtral Small 24B 2507	4.3s	$0.20	—
Codestral 2508	3.9s	$0.60	—
Mistral 7B Instruct	—	Free	—
Mistral Large 2407	21.8s	$4.00	—
Mistral Medium 3.1	11.5s	$1.20	—
Mixtral 8x22B Instruct	8.3s	$4.00	—
Mistral Large	22.3s	$4.00	—
Mistral Medium 3	10.7s	$1.20	—
Mistral Medium 3.5	—	$4.50	—
Mistral Nemo	6.3s	$0.02	—
Mistral Small 3.1 24B	17.6s	$0.45	—
Mistral Small Creative	7.5s	$0.20	—
Devstral Medium	5.0s	$1.20	—
Mistral Large 2411	15.7s	$4.00	—
Pixtral Large 2411	22.5s	$4.00	—
Devstral Small 1.1	3.0s	$0.20	—
Mistral Small 3.2 24B	14.1s	$0.14	—
Mistral Small 3	15.8s	$0.07	—
Mistral 7B Instruct v0.1	41.8s	$0.15	—
Ministral 8B	36.3s	$0.10	—
Ministral 3B	2.4s	$0.04	—
Pixtral 12B	7.6s	$0.10	—
Mistral 7B Instruct v0.3	3.7s	$0.20	—
Mistral Tiny	5.4s	$0.25	—
Mistral 7B Instruct v0.2	4.1s	$0.20	—
Mixtral 8x7B Instruct	9.0s	$0.54	—
Devstral 2 2512	13.7s	$1.20	—
Mistral Small 4	1.1s	$0.38	—