DevStral Small 2507: benchmark scores, price and ranking

DevStral Small 2507 from Mistral ranks #369 of 743 LLMs on the LLMs Tiger aggregate benchmark score (48.3), based on 1 benchmark results captured up to 2026-10-13.

Organisation: Mistral
Type: Open-weight / local model
Open weights: Yes
Released: 2025-07-10 (source: curated)
Aggregate rank: #369 of 743 (score 48.3)

Across the 1 benchmark where at least ten models are scored, DevStral Small 2507 is in the top 10% on 0 and in the top half on 0.

DevStral Small 2507 benchmark results

DevStral Small 2507 benchmark scores, ranked against all models measured on each benchmark
Benchmark	Score	Rank	Source	Captured
SWE-bench Verified	38.0%	#56 of 73	SWE-bench Verified leaderboard	2026-10-13

About DevStral Small 2507

DevStral Small 2507 is an open-weight model that can run on your own hardware from Mistral. It was released on 2025-07-10, according to the developer's announcement. No API price is listed for it yet. We hold 1 benchmark result for it across 1 area, and none is in the top tenth of its board.

Where DevStral Small 2507 stands

On our aggregate score, DevStral Small 2507 is #369 of 743 ranked models, around the middle of the table. Its score is 48.3; the scale is centred on 50, so it sits 1.7 points below the typical model. That beats 50% of the ranked field. Among the 2 ranked Mistral models it is #2. The next one up is DevStral Small 2505 at 49.1.

DevStral Small 2507, benchmark by benchmark

DevStral Small 2507 lands below the middle on SWE-bench Verified: 38.0%, #56 of 73, with 75% of the field scoring higher. Claude Opus 4.7 (max) leads at 83.5%; the closest result above is GPT-4.1 at 39.6%. It covers real GitHub issues resolved. The board median is 56.4%, so DevStral Small 2507 is 18.4 points below it. Source: SWE-bench Verified leaderboard, captured 2026-10-13.

Release history

The Mistral model released before it was DevStral Small 2505 on 2025-05-21, which scores 49.1.

Reading these numbers

The 1 score on this page comes from 1 source: SWE-bench Verified leaderboard (1). 1 of them come from an official leaderboard and 0 from datasets or third-party runs. With 1 results the aggregate rank can move a lot when one new score arrives.

Models ranked nearby

#366 Ring-2.6-1T (48.3)
#367 Qwen2.5 Coder 32B Instruct (48.3)
#368 mini-SWE-agent + Gemini 3.8 Flash (48.3)
#370 ernie-5.1 (48.3)
#371 Opus 4.5 (Thinking, 8K) (48.2)
#372 qwen3.5-35b-a3b (48.2)