MMLU#8 of 28
acc
View run →77.6%±0.3%
Evaluated across 13 benchmarks. Ranks in the top 3 on 0 of 13. Strongest showing on MMLU (77.6% acc, #8 of 28). Weakest on MBPP(Instruct) (0.0% pass@1, #21 of 24).
| Benchmark | Metric | Score | Rank | Actions |
|---|---|---|---|---|
| MMLU | acc | 77.6%±0.3% | #8 of 28 | View run → |
| MMLU-Pro | exact_match | 29.4%±0.4% | #24 of 28 | View run → |
| GPQA Diamond | acc | 27.8%±3.2% | #21 of 28 | View run → |
| GPQA Extended | acc | 26.9%±1.9% | #17 of 28 | View run → |
| GPQA Main | acc | 26.1%±2.1% | #21 of 28 | View run → |
| MGSM | exact_match | 23.5%±0.6% | #28 of 28 | View run → |
| IFEval | prompt_level_strict_acc | 23.3%±1.8% | #28 of 28 | View run → |
| LongBench | aggregate | 10.2%±0.2% | #14 of 17 | View run → |
| GSM8K | exact_match | 9.2%±0.8% | #25 of 28 | View run → |
| EQ-Bench | eqbench | 3.6±1.4 | #26 of 28 | View run → |
| BBH | exact_match | 3.3%±0.2% | #18 of 28 | View run → |
| MBPP | pass@1 | 0.4%±0.3% | #22 of 28 | View run → |
| MBPP(Instruct) | pass@1 | 0.0%±0.0% | #21 of 24 | View run → |
Citation
FrozeBench. "microsoft/phi-4-reasoning-plus." https://frozebench.com/models/microsoft-phi-4-reasoning-plus. Retrieved 2026-06-04.
BibTeX
@misc{frozebench_microsoft_phi_4_reasoning_plus,
title = {microsoft/phi-4-reasoning-plus},
howpublished = {\url{https://frozebench.com/models/microsoft-phi-4-reasoning-plus}},
year = {2026},
note = {FrozeBench. Retrieved 2026-06-04.}
}URL
https://frozebench.com/models/microsoft-phi-4-reasoning-plus