All tasks / Models

Devstral Small 2 (2512)

24B open-weights Devstral coding model under Apache 2.0; Mistral says it runs on a single RTX 4090.

  • Mistral AI
  • Open weights
  • Released 9 Dec 2025

Official page

Where it ranks

  • –
    Coding
    Models · Local · no result on this task’s benchmarks yet

Facts

Vendor
Mistral AI
Family
Devstral Small
Released
9 Dec 2025
Weights
Open: download and run it yourself
License
Apache-2.0
Parameters
24B
GPU memory to run locally
~15 GB (4-bit ≈ 24 B × 0.57 GB + 1.5 GB ≈ 15 GB, estimate; Mistral: runs on one RTX 4090 or a 32 GB Mac.)
Access
Free (open weights)
Last verified
18 Oct 2026

Price from openrouter.ai, 18 Oct 2026.

Benchmark results

BenchmarkResultEntry on the leaderboardAs ofSource
SWE-bench Verified (vendor-reported)68%Devstral Small 2
vendor-reported in model card, harness: not stated
18 Oct 2026huggingface.co

Where you can use it

Earlier versions

  • Devstral Small 1.1 (2507) · —

Sources