Microsoft: phi-3-medium-4k-instruct
1 benchmarks
Leaderboard position
#601
Overall rank (of 1122 models)
870
Elo rating
1
Benchmarks
Top 53.6%
Percentile
In Microsoft (of 19 models)
Above: Mistral AI: mistral-small-3.2-24b-instruct-2506 #600
Below: JetBrains: Mellum2 Thinking #602
Benchmark highlights
Nearest competitor from another company
The model from another company with the most similar benchmark scores.
appeared as nearest in 1 of 1 benchmarks
1/1
100.0%
Benchmarks
Predicted scores
Estimated from 25 similar models and the global Elo rating. These are projections, not measured results.
| Benchmark | Predicted | Confidence | Based on |
|---|---|---|---|
| Artificial Analysis Intelligence Index v5.0.1 | ~12 | 50% | — |
| MATH-500 | ~86.68 | 50% | — |
| AIME 2025 | ~66.69 | 50% | — |
| AIME 2026 | ~73.85 | 50% | — |
| HMMT Feb 2026 | ~65.35 | 50% | — |
| GPQA Diamond | ~58.47 | 50% | — |
| MMLU-Pro | ~61.66 | 50% | — |
| Humanity's Last Exam (no tools) | ~11.9 | 50% | — |
| IFEval | ~82.51 | 50% | — |
| IFBench | ~53.63 | 50% | — |
| HumanEval | ~74.2 | 50% | — |
| MBPP | ~77.69 | 50% | — |
| Terminal-Bench v2.1 | ~21.82 | 50% | — |
| MMLU | ~68.12 | 50% | — |
| SWE-bench Verified | ~39.79 | 50% | — |
| SWE-bench Pro | ~18.68 | 50% | — |
| SciCode | ~27.11 | 50% | — |
| AA-LCR | ~43.38 | 50% | — |
| Humanity's Last Exam (with tools) | ~16.05 | 50% | — |
| -(CoderPPL) | ~-1.69 | 50% | — |
| LiveCodeBench v6 | ~56.33 | 50% | — |
| MMMU-Pro | ~53.25 | 50% | — |
| MATH | ~61.49 | 50% | — |
| BBH | ~65.72 | 50% | — |
| ARC-Easy | ~85.5 | 50% | — |