BenchLeader
Menu
LeaderboardsModelsProvidersBenchmarksCompareReleasesBlogMethodology
MicrosoftAuto-detected

Wizardlm 70B

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. Reasoning31
  2. Coding30
  3. Human preference32

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 Prompts1172#304LMArena

Coding

BenchmarkdefaultSource
LMArena Coding1193#305LMArena

Human preference

BenchmarkdefaultSource
LMArena Text1184#292LMArena

See also

All Microsoft models

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