Data sourced primarily from official releases (GitHub, Hugging Face, papers), then benchmark leaderboards, then third-party evaluators. Learn about our data methodology

Devstral Small 1.0

Model basics

Reasoning traces

Not supported

Thinking modes

Thinking modes not supported

Context length

128K tokens

Max output length

4096 tokens

Model type

编程大模型

Release date

2025-05-26

Model file size

48GB

MoE architecture

Total params / Active params

240.0B / N/A

Knowledge cutoff

No data

Devstral Small 1.0

Open source & experience

Code license

Apache 2.0

Weights license

Apache 2.0- 免费商用授权

GitHub repo

GitHub link unavailable

Hugging Face

https://huggingface.co/mistralai/Devstral-Small-2505

Live demo

https://chat.mistral.ai/chat

Devstral Small 1.0

Official resources

Paper

Upgrading agentic coding capabilities with the new Devstral models

DataLearnerAI blog

No blog post yet

Devstral Small 1.0

API details

API speed

3/5

💡Default unit: $/1M tokens. If vendors use other units, follow their published pricing.

Standard pricingStandard

Modality	Input	Output
Text	$0.1	$0.3

Devstral Small 1.0

Benchmark Results

Devstral Small 1.0 currently shows benchmark results led by SWE-bench Verified (81 / 93, score 46.80). This page also consolidates core specs, context limits, and API pricing so you can evaluate the model from benchmark results and deployment constraints together.

Thinking

All modes Normal