Skip to content
RankShark

Devstral 2 2512

Devstral 2 is a state-of-the-art open-source model by Mistral AI specializing in agentic coding.

ModelOpen sourceby Mistralsince 2025
29.9

SharkScore

#224 in Models

Why it scores 29.9

Devstral 2 2512 has a SharkScore of 29.9/100, ranked #224 of 607 in Models (updated 2026-11-19).

Each part compares Devstral 2 2512 with other listings in Models. Evidence covers 87% of the score's weight.

Adoption
0 · 30%
User reviews
no data · 25%
Maintenance
27 · 20%
Trust & transparency
80 · 15%
Momentum
no data · 10%

Paying never changes this score. How ranking works · method v2026-11-18.2

RankShark review of Devstral 2 2512

Automated evaluation by RankShark · 18 Nov 2026 · method v1.0 · How we review

We were not paid for this review.

As of 2026-11-18, Devstral 2 2512 is a 123B-parameter coding model with a 262,144-token OpenRouter context window and usage-based token pricing. (E1, E2)

Devstral 2 2512 is Mistral's instruction-tuned model for software engineering and agentic coding. The vendor describes use for exploring codebases, editing multiple files, and building software engineering agents. (E1)

Best for

  • Developers building coding agents or AI code assistants. · model card
  • Teams that want to serve the model through an API or run it with supported local-serving frameworks. · model card, OpenRouter

Not for

  • Workflows that require image input; the listing specifies text and file input, with text output. · OpenRouter

Key features

  • Agentic coding. The vendor says the model is designed for coding agents, codebase exploration, and editing multiple files. · model card
  • Context window. The OpenRouter listing reports a 262,144-token context window. · OpenRouter
  • Tool calling. The vendor says Devstral 2 supports tool calling, with examples for vLLM and SGLang. · model card
  • Model size and format. The vendor describes this as a 123B-parameter instruction model in FP8. · model card
  • Local serving options. The vendor documents vLLM, SGLang, Transformers, llama.cpp, and Ollama as ways to run the model. · model card

Pros

  • The listing reports a context window of 262,144 tokens. · OpenRouter
  • The vendor documents both API use and local-serving options. · model card
  • The vendor documents tool-calling examples for vLLM and SGLang. · model card

Cons

  • The vendor advises using its API if local serving gives subpar performance. · model card
  • No free plan or free trial stated on the pages we checked.
  • Whether customer data is used for model training is not stated on the pages we checked.

Pricing

PlanPriceLimits
OpenRouter usage-based model pricing$0.40 per 1M input tokens; $2 per 1M output tokens262,144-token context window

Checked 18 Nov 2026

Free plan
Not stated (checked 18 Nov 2026)
Free trial
Not stated (checked 18 Nov 2026)
Trains on customer data
Not stated (checked 18 Nov 2026)
Data retention
Not stated (checked 18 Nov 2026)
Data regions
Not stated (checked 18 Nov 2026)
Data processing agreement
Not stated (checked 18 Nov 2026)
AI models used
Devstral 2 123B Instruct 2512 · model card, OpenRouter
API
OpenRouter lists chat completions, completions, messages, and responses endpoints. · OpenRouter
SDKs
Not stated (checked 18 Nov 2026)
Webhooks
Not stated (checked 18 Nov 2026)
Platforms
Local-serving frameworks documented by the vendor: vLLM, SGLang, Transformers, llama.cpp, and Ollama. · model card
Security claims
No security certification stated (checked 18 Nov 2026)

What we measured

Context window (OpenRouter listing) · now
262,144 tokens
Price per million tokens (OpenRouter listing) · now
$0.4 in, $2 out
Package downloads · 30 days
7,075
How we tested · sources
  • Hourly uptime and response-time checks of the product's URL.
  • Repository, release and download data from GitHub, npm, PyPI and Hugging Face where they exist.
  • We did not sign up, pay, or use the product as a customer.
  • 18 Nov 2026: Plans or prices changed on the pricing page.

Something wrong? Report an error with a public link that shows it; corrections re-run the review.

Top alternatives to Devstral 2 2512

See all
  • 67.1

    Qwen3.8 27B is an open-weight dense vision-language model from Qwen.

    ModelOpen source
  • 65.5

    Qwen3.6-35B-A3B is an open-weight multimodal model from Alibaba Cloud with 35 billion total parameters and 3 billion active parameters per t

    ModelOpen source
  • 64.7

    Qwen3.8 Flash is a multimodal reasoning model from Alibaba.

    ModelOpen source
  • 64.3

    Qwen3.6 27B is a dense 27-billion-parameter language model from the Qwen Team at Alibaba, released in April 2026.

    ModelOpen source