BenchLeader
Menu
LeaderboardsModelsProvidersBenchmarksCompareReleasesBlogMethodology
deepseekOpen weightsAuto-detected

DeepSeek V3.2 Reasoner

Not yet ranked: too few independent results so far. Released 1 Dec 2025.

Blended price
–
Output speed
–
First answer
–
Context
–
How it scores by categoryDashed line = average model (50). One step of 15 = one standard deviation.
  1. Coding54

Benchmark results

One column per reasoning effort. Rank is among every configuration of every model on that benchmark. Hover a score for the run it came from.

Coding

BenchmarkdefaultSource
SWE-bench Verified (bash only)60.0%#23SWE-bench
SWE-bench Verified (any scaffold)not in index60.0%#31SWE-bench

See also

All deepseek models

Data as of 9 Sept 2026. Compare with another model.