Model comparison
GPT-5 Nano vs Phi-4 Mini
GPT-5 Nano is the stronger model overall, scoring 33.4 to 30.4 on the Noometry Index.
Last verified . 1 shared benchmarks.
MicrosoftSummary
- They share 1 benchmark with published results for both. GPT-5 Nano scores higher in 2 categories and Phi-4 Mini in 1 category; 3 gaps are clear of the uncertainty.
- The widest gap is in knowledge, where GPT-5 Nano leads 35.8 to 24.8.
- The biggest single-benchmark swing is Vectara Hallucination Rate: 10.5% for GPT-5 Nano and 23.5% for Phi-4 Mini.
- Both cost about the same: $0.05 input and $0.40 output per million tokens.
- GPT-5 Nano accepts more context: 400K tokens versus 128K.
- Phi-4 Mini has downloadable open weights; the other is API-only.
Side by side
| GPT-5 Nano | Phi-4 Mini | |
|---|---|---|
| Provider | OpenAI | Microsoft |
| Noometry Index | 33.4 | 30.4 |
| Released | 2025-08-07 | 2024-12-11 |
| Weights | Proprietary | Open |
| Context window | 400K | 128K |
| Max output | 128K | 4K |
| Input $ / M tokens | $0.05 | $0.075 |
| Output $ / M tokens | $0.40 | $0.30 |
| Results tracked | 48 | 3 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding GPT-5 Nano leads
GPT-5 Nano: 33.3 (#248), Phi-4 Mini: 27.6 (#312)
| Benchmark | GPT-5 Nano | Phi-4 Mini |
|---|---|---|
| SWE-bench Verified (bash only) | 34.8% | — |
| SciCode | — | 10.8% |
| WeirdML | 38.1% | — |
| LMArena Coding | 1351 | — |
| ALE-Bench | 718.67 | — |
Agentic & Tool Use Not comparable
GPT-5 Nano: 25.6 (#99), Phi-4 Mini: —
| Benchmark | GPT-5 Nano | Phi-4 Mini |
|---|---|---|
| Terminal-Bench | 21.8% | — |
| Berkeley Function Calling Leaderboard | 51.5% | — |
Reasoning Phi-4 Mini leads
GPT-5 Nano: 16.4 (#291), Phi-4 Mini: 21.7 (#190)
| Benchmark | GPT-5 Nano | Phi-4 Mini |
|---|---|---|
| ARC-AGI-2 | 2.6% | — |
| Kagi LLM Benchmark | 62.2% | — |
| ARC-AGI-1 | 20.7% | — |
| CritPt | — | 0% |
| Chess Puzzles | 27% | — |
| LMArena Hard Prompts | 1329 | — |
| Mystery Game Puzzles | 9% | — |
| DTBench | 62.7% | — |
| LMCA | 7.9% | — |
| Epoch Capabilities Index | 139.39 | — |
| ForecastBench | 59.1 | — |
Math Not comparable
GPT-5 Nano: 29.1 (#221), Phi-4 Mini: —
| Benchmark | GPT-5 Nano | Phi-4 Mini |
|---|---|---|
| FrontierMath (Tiers 1-3) | 20% | — |
| FrontierMath Tier 4 | 2.4% | — |
| OTIS Mock AIME 2024-2025 | 81.1% | — |
| ProofBench | 12% | — |
| Omni-MATH | 54.6% | — |
| LMArena Math | 1318 | — |
| MATH Level 5 | 95.2% | — |
| FrontierMath (Feb 2025 set) | 8.3% | — |
| FrontierMath Tier 4 (v1) | 2.1% | — |
Knowledge GPT-5 Nano leads
GPT-5 Nano: 35.8 (#166), Phi-4 Mini: 24.8 (#261)
| Benchmark | GPT-5 Nano | Phi-4 Mini |
|---|---|---|
| Vectara Hallucination Rate | 10.5% | 23.5% |
| GPQA Diamond | 69.4% | — |
| SimpleQA Verified | 11.7% | — |
| MMLU-Pro | 77.8% | — |
| GPQA (HELM) | 67.9% | — |
| LMArena Expert | 1321 | — |
Multimodal Not comparable
GPT-5 Nano: 30.7 (#102), Phi-4 Mini: —
| Benchmark | GPT-5 Nano | Phi-4 Mini |
|---|---|---|
| LMArena Vision | 1159 | — |
| VPCT | 37.2% | — |
Multilingual Not comparable
GPT-5 Nano: 45.5 (#167), Phi-4 Mini: —
| Benchmark | GPT-5 Nano | Phi-4 Mini |
|---|---|---|
| LMArena Non-English | 1313 | — |
| LMArena Chinese | 1355 | — |
| LMArena German | 1329 | — |
| LMArena Korean | 1269 | — |
| LMArena Russian | 1296 | — |
| LMArena Spanish | 1360 | — |
Instruction Following Not comparable
GPT-5 Nano: 75.2 (#72), Phi-4 Mini: —
| Benchmark | GPT-5 Nano | Phi-4 Mini |
|---|---|---|
| IFEval | 93.2% | — |
| LMArena Instruction Following | 1306 | — |
Long Context Not comparable
GPT-5 Nano: 31.3 (#275), Phi-4 Mini: —
| Benchmark | GPT-5 Nano | Phi-4 Mini |
|---|---|---|
| Fiction.LiveBench | 44.4% | — |
| LMArena Longer Query | 1313 | — |
Writing & Preference Not comparable
GPT-5 Nano: 39.4 (#244), Phi-4 Mini: —
| Benchmark | GPT-5 Nano | Phi-4 Mini |
|---|---|---|
| LMArena Text | 1320 | — |
| LMArena Creative Writing | 1250 | — |
| EQ-Bench Creative Writing | 705 | — |
| WildBench | 80.6% | — |
| LMArena Multi-Turn | 1311 | — |
Frequently asked questions
Is GPT-5 Nano better than Phi-4 Mini?
GPT-5 Nano is the stronger model overall, scoring 33.4 to 30.4 on the Noometry Index.
Which is cheaper, GPT-5 Nano 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; GPT-5 Nano lists at $0.05 and $0.40.
Is GPT-5 Nano or Phi-4 Mini better for coding?
GPT-5 Nano scores higher on coding benchmarks: 33.3 versus 27.6 in the Noometry coding category.
Which has the bigger context window?
GPT-5 Nano does, with 400K tokens against 128K.
How many benchmarks do GPT-5 Nano and Phi-4 Mini share?
1 benchmark has published results for both models. GPT-5 Nano has 48 scored results on Noometry and Phi-4 Mini has 3.