BenchLeader
Menu
LeaderboardsModelsProvidersBenchmarksCompareReleasesBlogMethodology
AlibabaAuto-detected

Qwen3 5

Not yet ranked: too few independent results so far.

Blended price
–
Output speed
–
First answer
–
Context
–
How it scores by categoryDashed line = average model (50). One step of 15 = one standard deviation.
  1. Reasoning67
  2. Coding66
  3. Human preference66

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.

Reasoning

BenchmarkdefaultSource
LMArena Hard Prompts1492#37LMArena

Coding

BenchmarkdefaultSource
LMArena Coding1514#39LMArena

Human preference

BenchmarkdefaultSource
LMArena Text1466#45LMArena

See also

All Alibaba models

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