BenchLeader
Menu
LeaderboardsModelsProvidersBenchmarksCompareReleasesBlogMethodology
MicrosoftAuto-detected

Phi 3 Mini

Best configuration ranks #572 of 610 on the BenchLeader Index at 36.8 ±2.9. Last measured 2 Sept 2026.

Blended price
–
Output speed
–
First answer
–
Context
4k
How it scores by categoryDashed line = average model (50). One step of 15 = one standard deviation.
  1. Overall index37
  2. Reasoning29
  3. Coding29
  4. Agents & tools34
  5. Instruction following29
  6. Human preference26
  7. Composite36

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 Prompts1153#315LMArena
GPQA Diamond (AA)not in index31.9%#483Artificial Analysis
Humanity's Last Exam (AA)not in index5.0%#382Artificial Analysis

Coding

BenchmarkdefaultSource
LMArena Coding1187#307LMArena

Agents & tools

BenchmarkdefaultSource
Terminal-Bench Hard0.0%#345Artificial Analysis
τ²-Bench Telecom (AA)not in index0.0%#375Artificial Analysis

Instruction following

BenchmarkdefaultSource
IFBench23.9%#377Artificial Analysis

Human preference

BenchmarkdefaultSource
LMArena Text1129#325LMArena

Composite

BenchmarkdefaultSource
AA Intelligence Index5.8#469Artificial Analysis

See also

All Microsoft models

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