Azure · model passport
Phi-4 Mini Instruct
phi-4-mini-instruct- Best percentile
- 14.9LiveCodeBench (Artificial Analysis)
- Composite standing
- 26.4shrunk · 4/27 axes
What the data says
Benchmarked on 4 of 27 tracked axes. Every figure below is the model's own published result on that board.
Weakest axes
- 1.9 pct — HHEM Hallucination (Vectara) (76.5%)
- 4 pct — IFBench (Artificial Analysis) (21.1%)
- 14.2 pct — MMLU-Pro (Artificial Analysis) (46.5%)
Uneven: 13 percentile points between its best and worst axis — 14.9 pct on LiveCodeBench (Artificial Analysis) versus 1.9 pct on HHEM Hallucination (Vectara).
- LiveCodeBench (Artificial Analysis)14.9 pct12.6%Verify →
- MMLU-Pro (Artificial Analysis)14.2 pct46.5%Verify →
- IFBench (Artificial Analysis)4 pct21.1%Verify →
- HHEM Hallucination (Vectara)1.9 pct76.5%Verify →
- OverallNot benchmarked
- WebDev ArenaNot benchmarked
- Vision ArenaNot benchmarked
- Factuality ArenaNot benchmarked
- Search ArenaNot benchmarked
- Document ArenaNot benchmarked
- Agent ArenaNot benchmarked
- GPQA DiamondNot benchmarked
- AIME (math)Not benchmarked
- FrontierMathNot benchmarked
- SimpleQA VerifiedNot benchmarked
- ARC-AGI-2Not benchmarked
- Humanity's Last ExamNot benchmarked
- SimpleBenchNot benchmarked
- LiveBenchNot benchmarked
- BFCL v4 (function calling)Not benchmarked
- SWE-bench Verified (live board)Not benchmarked
- Terminal-Bench 4.0Not benchmarked
- SWE-bench VerifiedNot benchmarked
- Terminal-BenchNot benchmarked
- CodingNot benchmarked
- Instruction followingNot benchmarked
- Hard promptsNot benchmarked
Percentiles are the model's standing within each board's own population. Axes marked "not benchmarked" have no published result for this model — that is an absence of data, not a low score.
Phi-4 Mini Instruct stands at 26.4 (raw mean 8.8) — measured on 4 of 27 axes.
Every claim is a computed fact from the current board data, phrased by a low-temperature model call and rejected by a validation gate if any number is not in the source row. How the leaderboards work →