mistralai
Mistral: Devstral 2 2512
mistralai/devstral-2512
Our take
AI-assisted analysis based on the sources below. Suitability is an editorial assessment; test it on your own workload.
Mistral's Devstral 2 2512 (mistralai/devstral-2512) is a 123B-parameter dense instruct model in FP8 aimed at agentic software engineering, with a 256K-token context window and text-plus-file inputs, per the OpenRouter catalog and Mistral's model card. Mistral reports strong SWE-bench and terminal-benchmark results and improved generalization over earlier Devstral releases; these are vendor claims, while OpenRouter lists only partial third-party index data. It fits coding agents and assistants; validate on your own workloads before adopting.[1][2]
Strengths
- Mistral positions it for agentic coding: tool use, codebase exploration and multi-file edits.[1][2]
- The 256K-token context and text-plus-file inputs suit large repositories and long sessions.[1][2]
- Vendor-reported SWE-bench and terminal-benchmark results are described as improved over previous Devstral releases.[2]
Limitations
- Evidence gaps: knowledge cutoff and reasoning support are unspecified, and OpenRouter's third-party index coverage is incomplete; validate on your own workloads.[1]
- Benchmark figures are vendor-reported; independent verification is limited in this source pack.[1][2]
- Hugging Face lists the license as other, so review terms before production use; the vendor also notes local serving may underperform the hosted API.[2]
Where to use it
- Likely fit for agentic coding assistants and CLI agents; Mistral documents Mistral Vibe, Cline, OpenHands, SWE Agent and similar scaffolds.[2]
- May suit multi-file refactors and issue-driven fixes in large codebases given the long context.[1][2]
- Can run via the Mistral API or locally with vLLM, SGLang, transformers or community builds; compare local quality against the API first.[2]
Sources
- [1] OpenRouter catalog · Mistral: Devstral 2 2512Sources checked: Nov 13, 2026
- [2] Model documentation · Mistral: Devstral 2 2512Sources checked: Nov 13, 2026
About this model
Description below is the original text supplied by OpenRouter, shown in the source language.
Devstral 2 is a state-of-the-art open-source model by Mistral AI specializing in agentic coding. It is a 123B-parameter dense transformer model supporting a 256K context window. Devstral 2 supports exploring...
Specifications
- Model ID
- mistralai/devstral-2512
- Creator
- mistralai
- Added to OpenRouter
- 09.12.2025
- Context length
- 262,144
- Modality
- text+file->text
- Input modalities
- text, file
- Output modalities
- text
- Token pricing
- USD per 1M tokens: in $0.400 / out $2
- Knowledge cutoff
- —
- Expiration date
- —
Tasks and capabilities
A grade appears only where the source supplies a matching measurement for that task, the model is active and the cohort holds at least 10 comparable models. The capability column lists facts the source declares — supported formats, reasoning and tool parameters — which are not measured quality.
| Task | Grade | Source-declared capability |
|---|---|---|
| General quality | No measurement | — |
| Coding | CC — below the B band | — |
| Agents | CC — below the B band | parameters: structured_outputs, tool_choice, tools |
| Writing | No measurement | output: text |
| Reasoning | No measurement | — |
| Document analysis | No measurement | input: file |
| Vision | No measurement | — |
| Image generation | No measurement | — |
| Audio | No measurement | — |
| Video | No measurement | — |
| Embeddings | No measurement | — |
Evaluations
| Source | Benchmark | Value | Grade | Data retrieved |
|---|---|---|---|---|
| Artificial Analysis | Coding index | 31.3 | CC — below the B band | 13.11.2026 |
| Artificial Analysis | Agentic index | 2.3 | CC — below the B band | 13.11.2026 |
How grades work
Each letter grade compares a model only against other models scored by the same source and metric in the current snapshot. S means the top 10%, A the next 20%, B the next 30%, C the rest. Models without a published measurement are shown as not rated — a supported format alone never implies measured quality.
- S — S — top 10% of the cohort
- A — A — next 20% of the cohort
- B — B — next 30% of the cohort
- C — C — below the B band
Providers
Optional live provider statistics from OpenRouter. Missing values are unknown, not zero.
| Provider | Context length | Price in /1M | Price out /1M | Latency p50 (30m, s) | Throughput p50 (30m, tok/s) | Uptime (30m) |
|---|---|---|---|---|---|---|
| Mistral | 262,144 | $0.400 | $2 | — | — | — |
Provider data retrieved: 13.11.2026
Related articles
No published articles identify this model yet.