Model comparison
o4-mini vs Phi-4 Mini
o4-mini is the stronger model overall, scoring 41.6 to 30.4 on the Noometry Index. Phi-4 Mini costs 15× less per token, which makes it the better buy when o4-mini's lead doesn't matter for your workload.
Last verified . 2 shared benchmarks.
MicrosoftSummary
- They share 2 benchmarks with published results for both. o4-mini scores higher in 3 categories and Phi-4 Mini in 0 categories; 3 gaps are clear of the uncertainty.
- The widest gap is in knowledge, where o4-mini leads 43.5 to 24.8.
- Phi-4 Mini is cheaper at $0.075 / $0.30 per million input/output tokens, against $1.10 / $4.40 for o4-mini.
- o4-mini accepts more context: 200K tokens versus 128K.
- Phi-4 Mini has downloadable open weights; the other is API-only.
Side by side
| o4-mini | Phi-4 Mini | |
|---|---|---|
| Provider | OpenAI | Microsoft |
| Noometry Index | 41.6 | 30.4 |
| Released | 2025-04-16 | 2024-12-11 |
| Weights | Proprietary | Open |
| Context window | 200K | 128K |
| Max output | 100K | 4K |
| Input $ / M tokens | $1.10 | $0.075 |
| Output $ / M tokens | $4.40 | $0.30 |
| Results tracked | 60 | 3 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding o4-mini leads
o4-mini: 41.6 (#110), Phi-4 Mini: 27.6 (#312)
| Benchmark | o4-mini | Phi-4 Mini |
|---|---|---|
| SWE-bench Verified (bash only) | 45% | — |
| Aider Polyglot | 72% | — |
| SciCode | — | 10.8% |
| GSO | 3.6% | — |
| WeirdML | 52.6% | — |
| LMArena Coding | 1369 | — |
| CadEval | 62% | — |
| ALE-Bench | 826.17 | — |
| AlgoTune | 1.72 | — |
Agentic & Tool Use Not comparable
o4-mini: 32.3 (#57), Phi-4 Mini: —
| Benchmark | o4-mini | Phi-4 Mini |
|---|---|---|
| Berkeley Function Calling Leaderboard | 53.2% | — |
| GDPval | 25.3% | — |
| METR Time Horizons | 63.9% | — |
Reasoning o4-mini leads
o4-mini: 24.6 (#143), Phi-4 Mini: 21.7 (#190)
| Benchmark | o4-mini | Phi-4 Mini |
|---|---|---|
| CritPt | 0.6% | 0% |
| ARC-AGI-2 | 6.1% | — |
| SimpleBench | 38.7% | — |
| Kagi LLM Benchmark | 67.6% | — |
| ARC-AGI-1 | 58.7% | — |
| Chess Puzzles | 26% | — |
| EnigmaEval | 9.2% | — |
| LMArena Hard Prompts | 1351 | — |
| Mystery Game Puzzles | 5% | — |
| DTBench | 77.6% | — |
| LMCA | 26.5% | — |
| Epoch Capabilities Index | 145.63 | — |
| ForecastBench | 61.8 | — |
Math Not comparable
o4-mini: 40.2 (#79), Phi-4 Mini: —
| Benchmark | o4-mini | Phi-4 Mini |
|---|---|---|
| FrontierMath (Tiers 1-3) | 36.1% | — |
| FrontierMath Tier 4 | 4.9% | — |
| OTIS Mock AIME 2024-2025 | 81.7% | — |
| Omni-MATH | 72% | — |
| LMArena Math | 1390 | — |
| MATH Level 5 | 97.8% | — |
| FrontierMath (Feb 2025 set) | 24.8% | — |
| FrontierMath Tier 4 (v1) | 6.3% | — |
Knowledge o4-mini leads
o4-mini: 43.5 (#83), Phi-4 Mini: 24.8 (#261)
| Benchmark | o4-mini | Phi-4 Mini |
|---|---|---|
| Vectara Hallucination Rate | 18.6% | 23.5% |
| GPQA Diamond | 79.6% | — |
| Humanity's Last Exam | 18.1% | — |
| SimpleQA Verified | 19.6% | — |
| MMLU-Pro | 82% | — |
| Confabulations | 15.8% | — |
| GPQA (HELM) | 73.5% | — |
| LMArena Expert | 1344 | — |
Multimodal Not comparable
o4-mini: 39.5 (#46), Phi-4 Mini: —
| Benchmark | o4-mini | Phi-4 Mini |
|---|---|---|
| LMArena Vision | 1194 | — |
| GeoBench | 64% | — |
| VPCT | 57.5% | — |
Multilingual Not comparable
o4-mini: 47.2 (#149), Phi-4 Mini: —
| Benchmark | o4-mini | Phi-4 Mini |
|---|---|---|
| LMArena Non-English | 1336 | — |
| LMArena Chinese | 1355 | — |
| LMArena French | 1362 | — |
| LMArena German | 1336 | — |
| LMArena Japanese | 1308 | — |
| LMArena Korean | 1313 | — |
| LMArena Russian | 1334 | — |
| LMArena Spanish | 1346 | — |
Instruction Following Not comparable
o4-mini: 75.4 (#63), Phi-4 Mini: —
| Benchmark | o4-mini | Phi-4 Mini |
|---|---|---|
| IFEval | 92.8% | — |
| LMArena Instruction Following | 1321 | — |
Long Context Not comparable
o4-mini: 45.5 (#31), Phi-4 Mini: —
| Benchmark | o4-mini | Phi-4 Mini |
|---|---|---|
| Fiction.LiveBench | 77.8% | — |
| LMArena Longer Query | 1315 | — |
Writing & Preference Not comparable
o4-mini: 54.2 (#148), Phi-4 Mini: —
| Benchmark | o4-mini | Phi-4 Mini |
|---|---|---|
| LMArena Text | 1353 | — |
| LMArena Creative Writing | 1294 | — |
| Short-Story Creative Writing | 75% | — |
| WildBench | 85.4% | — |
| LMArena Multi-Turn | 1350 | — |
Frequently asked questions
Is o4-mini better than Phi-4 Mini?
o4-mini is the stronger model overall, scoring 41.6 to 30.4 on the Noometry Index. Phi-4 Mini costs 15× less per token, which makes it the better buy when o4-mini's lead doesn't matter for your workload.
Which is cheaper, o4-mini or Phi-4 Mini?
Phi-4 Mini is cheaper. It lists at $0.075 per million input tokens and $0.30 per million output tokens; o4-mini lists at $1.10 and $4.40.
Is o4-mini or Phi-4 Mini better for coding?
o4-mini scores higher on coding benchmarks: 41.6 versus 27.6 in the Noometry coding category.
Which has the bigger context window?
o4-mini does, with 200K tokens against 128K.
How many benchmarks do o4-mini and Phi-4 Mini share?
2 benchmarks have published results for both models. o4-mini has 60 scored results on Noometry and Phi-4 Mini has 3.