Devstral 2 2512: API price, context window and how to connect

Devstral 2 2512 by Mistral AI: $0.4 per million input tokens, $2 per million output tokens, a 262K-token window. What it is good for, what to compare it with and how to connect it in SKILLOK.

Updated Copy as Markdown
In this article

In short#

Devstral 2 is Mistral AI's open model for agentic coding: dense, 123 billion parameters, a 256-thousand-token window. It can explore a codebase and edit it step by step.

Specifications#

Provider Mistral AI
Released 2025-12-09
Context window 262K tokens
Accepts text, files
Answers with text
Reasoning mode no
Tool calling yes
Open weights yes
Identifier mistralai/devstral-2512

Price#

Per 1M tokens In dollars
Input (question, context) $0.4
Output (answer) $2

For reference: a conversation of 20 questions of ~500 tokens and answers of ~800 tokens costs about $0.036. Models in reasoning mode spend more tokens on an answer than the text shows: the reasoning is paid for too.

What it is good for#

  • A developer agent
  • Exploring unfamiliar code
  • An open model for code

Similar models#

Catalog models with the same kind of output, closest in the price of a typical request.

Model Input Output Window Price
Qwen3.8 27B $0.24 $2.2 1M about the same
Mistral Large 4 $0.68 $2.09 1.05M 1.2× more expensive
DeepSeek V4 Pro 0423 $0.955 $1.911 1.05M 1.3× more expensive
Qwen3.7 Plus $0.32 $1.28 1M 1.5× cheaper
Kimi K2.7 Code $0.671 $3.35 262K 1.7× more expensive

Other Mistral AI models#

How to connect in SKILLOK#

Devstral 2 2512 works through OpenRouter: add an OpenRouter key in Settings → AI and pick the mistralai/devstral-2512 model. It then becomes available in Space AI, in tasks and to the repository agent. In detail — Connecting a model in SKILLOK.

Where the data comes from#

Prices, the context window and capabilities come from the public OpenRouter catalog as of 2026-10-19. A provider may change the price — the exact figure is always on the model's OpenRouter page.

Was this helpful?