Sulat.com
AI models
$10 off the fastest DeepSeek V4.1 Flash, Kimi K3 and GLM 5.3 from Synthetic
Merge Gateway logo

Model details

Devstral 2

Devstral the listed price is a 123-billion-parameter model positioned as a workhorse for agentic software engineering, with a description that emphasizes tool use to explore codebases, edit multiple files, and power software-engineering agents. This framing is reinforced by independent third-party coverage published the day after launch that frames the release squarely around agentic coding workflows. The model's identity as "devstral-2" is consistent across the catalog entry and its open distribution page, where it sits inside a larger family of Devstral releases aimed at developer-facing tasks rather than general chat.

In practice, Devstral the listed price is designed to slot directly into existing agentic coding harnesses, with documented support for Claude Code, OpenCode, Hermes Agent, and OpenClaw through local runtime launch commands. The open distribution listing reports substantial early traction, recording hundreds of thousands of downloads, suggesting that developers are actively experimenting with it as a backend for autonomous code workflows. The combination of large parameter count, tool-calling orientation, and broad harness compatibility makes it a practical fit for teams building multi-step coding agents that need to navigate, edit, and reason across complex repositories.

Merge Gatewaymistral/devstral-2512devstral

Quick Info

Powered by
Provider
Merge Gateway
Model key
mistral/devstral-2512
Release date
Dec 9, 2025
Last updated
Dec 9, 2025
Knowledge cutoff
2025-12
Input modalities
Output modalities
Capabilities

Cost

Input token cost
$0.40
Output token cost
$2.00

Limits

Output tokens
256,000 tokens
Context window
256,000 tokens

Transparent token rates

Compare Devstral 2 pricing

Rates are shown per one million tokens. Combined means one million input plus one million output tokens.

Browse this family

Latest news about Devstral 2

Merge Gateway

CoverageBenchmark

Artificial Analysis published a model profile for Devstral 2, released by Mistral in December 2025 as an open-weights model. The profile reports an Intelligence Index score of 9, placing it well above average among comparable open-weight non-reasoning models (median: 7). Devstral 2 supports text input, outputs text, an The evaluation shows Devstral 2 generating 132 output tokens per second, faster than the average of 100 tokens per second across the 39-model comparison class. The model is described as highly concise, generating only 18M tokens for the full Intelligence Index evaluation versus a median of 26M. Pricing is listed as $0.

Videos about Devstral 2

More models around Devstral 2