BenchLeader
Menu
LeaderboardsModelsProvidersBenchmarksCompareReleasesBlogMethodology
AlibabaAuto-detected

Qwen1 5 4B

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. Reasoning22
  2. Coding23
  3. Human preference21

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 Prompts1094#336LMArena

Coding

BenchmarkdefaultSource
LMArena Coding1131#329LMArena

Human preference

BenchmarkdefaultSource
LMArena Text1090#336LMArena

See also

All Alibaba models

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