Model comparison
GPT-5 Mini vs phi-3-medium 14B
GPT-5 Mini is the stronger model overall, scoring 41.8 to 29.7 on the Noometry Index.
Last verified . 3 shared benchmarks.
MicrosoftSummary
- They share 3 benchmarks with published results for both. GPT-5 Mini scores higher in 3 categories and phi-3-medium 14B in 0 categories; 3 gaps are clear of the uncertainty.
- The widest gap is in knowledge, where GPT-5 Mini leads 45.6 to 9.1.
- The biggest single-benchmark swing is MATH Level 5: 97.8% for GPT-5 Mini and 17.6% for phi-3-medium 14B.
- phi-3-medium 14B has downloadable open weights; the other is API-only.
Side by side
| GPT-5 Mini | phi-3-medium 14B | |
|---|---|---|
| Provider | OpenAI | Microsoft |
| Noometry Index | 41.8 | 29.7 |
| Released | 2025-08-07 | 2024-04-23 |
| Weights | Proprietary | Open |
| Context window | 400K | — |
| Max output | 128K | — |
| Input $ / M tokens | $0.25 | — |
| Output $ / M tokens | $2 | — |
| Results tracked | 60 | 13 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding GPT-5 Mini leads
GPT-5 Mini: 40.1 (#146), phi-3-medium 14B: 36.8 (#201)
| Benchmark | GPT-5 Mini | phi-3-medium 14B |
|---|---|---|
| SWE-bench Verified | 64.7% | — |
| SWE-bench Verified (bash only) | 59.8% | — |
| SWE-bench Multilingual | 39.7% | — |
| SciCode | 39.2% | — |
| WeirdML | 52.7% | — |
| BigCodeBench Instruct | — | 37.6% |
| LMArena Coding | 1406 | — |
| BigCodeBench Complete | — | 48.7% |
| ALE-Bench | 799.77 | — |
| AlgoTune | 1.38 | — |
Agentic & Tool Use Not comparable
GPT-5 Mini: 31.1 (#70), phi-3-medium 14B: —
| Benchmark | GPT-5 Mini | phi-3-medium 14B |
|---|---|---|
| Terminal-Bench | 34.8% | — |
| Berkeley Function Calling Leaderboard | 55.5% | — |
| Vending-Bench 2 | -31.18 | — |
Reasoning Not comparable
GPT-5 Mini: 23.9 (#168), phi-3-medium 14B: —
| Benchmark | GPT-5 Mini | phi-3-medium 14B |
|---|---|---|
| Epoch Capabilities Index | 145.52 | 121.23 |
| ARC-AGI-2 | 4.4% | — |
| Kagi LLM Benchmark | 70.3% | — |
| ARC-AGI-1 | 54.3% | — |
| CritPt | 0% | — |
| Chess Puzzles | 30% | — |
| EnigmaEval | 8.2% | — |
| LMArena Hard Prompts | 1380 | — |
| Mystery Game Puzzles | 10% | — |
| DTBench | 80.5% | — |
| LMCA | 34.2% | — |
| Adversarial NLI | — | 55.8% |
| BIG-Bench Hard | — | 81.4% |
| ForecastBench | 61 | — |
| HellaSwag | — | 82.4% |
| WinoGrande | — | 81.5% |
Math GPT-5 Mini leads
GPT-5 Mini: 46.7 (#69), phi-3-medium 14B: 27.3 (#250)
| Benchmark | GPT-5 Mini | phi-3-medium 14B |
|---|---|---|
| MATH Level 5 | 97.8% | 17.6% |
| FrontierMath (Tiers 1-3) | 46.7% | — |
| FrontierMath Tier 4 | 12.2% | — |
| OTIS Mock AIME 2024-2025 | 86.7% | — |
| ProofBench | 9% | — |
| Omni-MATH | 72.2% | — |
| LMArena Math | 1378 | — |
| FrontierMath (Feb 2025 set) | 27.2% | — |
| FrontierMath Tier 4 (v1) | 6.3% | — |
Knowledge GPT-5 Mini leads
GPT-5 Mini: 45.6 (#86), phi-3-medium 14B: 9.1 (#306)
| Benchmark | GPT-5 Mini | phi-3-medium 14B |
|---|---|---|
| GPQA Diamond | 75% | 27.6% |
| Humanity's Last Exam | 19.4% | — |
| SimpleQA Verified | 21.6% | — |
| MMLU-Pro | 83.5% | — |
| Confabulations | 13.3% | — |
| Vectara Hallucination Rate | 12.9% | — |
| GPQA (HELM) | 75.6% | — |
| LMArena Expert | 1379 | — |
| ARC (AI2) Challenge | — | 91.6% |
| MMLU | — | 78% |
| OpenBookQA | — | 87.4% |
| TriviaQA | — | 73.9% |
Multimodal Not comparable
GPT-5 Mini: 35.6 (#85), phi-3-medium 14B: —
| Benchmark | GPT-5 Mini | phi-3-medium 14B |
|---|---|---|
| LMArena Vision | 1202 | — |
| VPCT | 40.2% | — |
Multilingual Not comparable
GPT-5 Mini: 48.9 (#137), phi-3-medium 14B: —
| Benchmark | GPT-5 Mini | phi-3-medium 14B |
|---|---|---|
| LMArena Non-English | 1363 | — |
| LMArena Chinese | 1385 | — |
| LMArena French | 1386 | — |
| LMArena German | 1366 | — |
| LMArena Japanese | 1341 | — |
| LMArena Korean | 1308 | — |
| LMArena Russian | 1362 | — |
| LMArena Spanish | 1355 | — |
Instruction Following Not comparable
GPT-5 Mini: 76.2 (#46), phi-3-medium 14B: —
| Benchmark | GPT-5 Mini | phi-3-medium 14B |
|---|---|---|
| IFEval | 92.7% | — |
| LMArena Instruction Following | 1357 | — |
Long Context Not comparable
GPT-5 Mini: 41.9 (#132), phi-3-medium 14B: —
| Benchmark | GPT-5 Mini | phi-3-medium 14B |
|---|---|---|
| Fiction.LiveBench | 69.4% | — |
| LMArena Longer Query | 1355 | — |
Writing & Preference Not comparable
GPT-5 Mini: 55.2 (#148), phi-3-medium 14B: —
| Benchmark | GPT-5 Mini | phi-3-medium 14B |
|---|---|---|
| LMArena Text | 1373 | — |
| LMArena Creative Writing | 1325 | — |
| Short-Story Creative Writing | 83.1% | — |
| EQ-Bench Creative Writing | 1313 | — |
| WildBench | 85.5% | — |
| LMArena Multi-Turn | 1363 | — |
Frequently asked questions
Is GPT-5 Mini better than phi-3-medium 14B?
GPT-5 Mini is the stronger model overall, scoring 41.8 to 29.7 on the Noometry Index.
Is GPT-5 Mini or phi-3-medium 14B better for coding?
GPT-5 Mini scores higher on coding benchmarks: 40.1 versus 36.8 in the Noometry coding category.
How many benchmarks do GPT-5 Mini and phi-3-medium 14B share?
3 benchmarks have published results for both models. GPT-5 Mini has 60 scored results on Noometry and phi-3-medium 14B has 13.