Model comparison
GPT-5.4 nano vs Phi 3 Mini 4k Instruct
GPT-5.4 nano is the stronger model overall, scoring 41.9 to 27.4 on the Noometry Index.
Last verified . 18 shared benchmarks.
MicrosoftSummary
- They share 18 benchmarks with published results for both. GPT-5.4 nano scores higher in 8 categories and Phi 3 Mini 4k Instruct in 0 categories; 8 gaps are clear of the uncertainty.
- The widest gap is in writing & preference, where GPT-5.4 nano leads 55.9 to 27.7.
- The biggest single-benchmark swing is Chess Puzzles: 30% for GPT-5.4 nano and 0% for Phi 3 Mini 4k Instruct.
- Phi 3 Mini 4k Instruct has downloadable open weights; the other is API-only.
Side by side
| GPT-5.4 nano | Phi 3 Mini 4k Instruct | |
|---|---|---|
| Provider | OpenAI | Microsoft |
| Noometry Index | 41.9 | 27.4 |
| Released | 2026-03-17 | 2024-04-23 |
| Weights | Proprietary | Open |
| Context window | 400K | — |
| Max output | 128K | — |
| Input $ / M tokens | $0.20 | — |
| Output $ / M tokens | $1.25 | — |
| Results tracked | 40 | 35 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding GPT-5.4 nano leads
GPT-5.4 nano: 43.1 (#82), Phi 3 Mini 4k Instruct: 26.1 (#319)
| Benchmark | GPT-5.4 nano | Phi 3 Mini 4k Instruct |
|---|---|---|
| LMArena Coding | 1404 | 1093 |
| SciCode | 46.9% | — |
| WeirdML | 49.2% | — |
| LiveBench Coding | — | 15.5% |
| ALE-Bench | 1,005 | — |
| HumanEval+ | — | 59.1% |
| MBPP+ | — | 54.2% |
Reasoning GPT-5.4 nano leads
GPT-5.4 nano: 23.8 (#159), Phi 3 Mini 4k Instruct: 13.6 (#326)
| Benchmark | GPT-5.4 nano | Phi 3 Mini 4k Instruct |
|---|---|---|
| Chess Puzzles | 30% | 0% |
| LMArena Hard Prompts | 1380 | 1072 |
| ARC-AGI-2 | 5.7% | — |
| Kagi LLM Benchmark | 39.7% | — |
| ARC-AGI-1 | 51.5% | — |
| CritPt | 9.3% | — |
| LiveBench Reasoning | — | 26.8% |
| Mystery Game Puzzles | 9% | — |
| DTBench | 80.3% | — |
| LiveBench Data Analysis | — | 34.7% |
| LMCA | 36.9% | — |
| Adversarial NLI | — | 52.8% |
| BIG-Bench Hard | — | 71.7% |
| Epoch Capabilities Index | 145.78 | — |
| ForecastBench | 57.3 | — |
| HellaSwag | — | 76.7% |
| LiveBench | — | 22.4% |
| WinoGrande | — | 70.8% |
Math GPT-5.4 nano leads
GPT-5.4 nano: 40.7 (#76), Phi 3 Mini 4k Instruct: 24.5 (#258)
| Benchmark | GPT-5.4 nano | Phi 3 Mini 4k Instruct |
|---|---|---|
| LMArena Math | 1406 | 1111 |
| FrontierMath (Tiers 1-3) | 44.9% | — |
| FrontierMath Tier 4 | 12.2% | — |
| OTIS Mock AIME 2024-2025 | 87.8% | — |
| ProofBench | 5% | — |
| LiveBench Math | — | 15.7% |
| FrontierMath (Feb 2025 set) | 25.9% | — |
| FrontierMath Tier 4 (v1) | 6.3% | — |
Knowledge GPT-5.4 nano leads
GPT-5.4 nano: 41.8 (#94), Phi 3 Mini 4k Instruct: 27.9 (#241)
| Benchmark | GPT-5.4 nano | Phi 3 Mini 4k Instruct |
|---|---|---|
| LMArena Expert | 1397 | 1045 |
| GPQA Diamond | 78.5% | — |
| SimpleQA Verified | 11.7% | — |
| Vectara Hallucination Rate | 3.1% | — |
| ARC (AI2) Challenge | — | 84.9% |
| MMLU | — | 68.8% |
| OpenBookQA | — | 88% |
| TriviaQA | — | 64% |
Multimodal Not comparable
GPT-5.4 nano: 36.0 (#73), Phi 3 Mini 4k Instruct: —
| Benchmark | GPT-5.4 nano | Phi 3 Mini 4k Instruct |
|---|---|---|
| LMArena Vision | 1196 | — |
Multilingual GPT-5.4 nano leads
GPT-5.4 nano: 48.8 (#136), Phi 3 Mini 4k Instruct: 26.5 (#275)
| Benchmark | GPT-5.4 nano | Phi 3 Mini 4k Instruct |
|---|---|---|
| LMArena Non-English | 1359 | 1021 |
| LMArena Chinese | 1391 | 1021 |
| LMArena French | 1396 | 1075 |
| LMArena German | 1368 | 1044 |
| LMArena Japanese | 1342 | 935 |
| LMArena Korean | 1321 | 905 |
| LMArena Russian | 1363 | 1022 |
| LMArena Spanish | 1368 | 1084 |
Instruction Following GPT-5.4 nano leads
GPT-5.4 nano: 72.1 (#139), Phi 3 Mini 4k Instruct: 48.0 (#298)
| Benchmark | GPT-5.4 nano | Phi 3 Mini 4k Instruct |
|---|---|---|
| LMArena Instruction Following | 1362 | 1053 |
| LiveBench Instruction Following | — | 39.1% |
Long Context GPT-5.4 nano leads
GPT-5.4 nano: 41.6 (#131), Phi 3 Mini 4k Instruct: 31.7 (#271)
| Benchmark | GPT-5.4 nano | Phi 3 Mini 4k Instruct |
|---|---|---|
| LMArena Longer Query | 1365 | 1044 |
Writing & Preference GPT-5.4 nano leads
GPT-5.4 nano: 55.9 (#137), Phi 3 Mini 4k Instruct: 27.7 (#295)
| Benchmark | GPT-5.4 nano | Phi 3 Mini 4k Instruct |
|---|---|---|
| LMArena Text | 1372 | 1073 |
| LMArena Creative Writing | 1313 | 1037 |
| LMArena Multi-Turn | 1382 | 1018 |
| LiveBench Language | — | 9.2% |
Frequently asked questions
Is GPT-5.4 nano better than Phi 3 Mini 4k Instruct?
GPT-5.4 nano is the stronger model overall, scoring 41.9 to 27.4 on the Noometry Index.
Is GPT-5.4 nano or Phi 3 Mini 4k Instruct better for coding?
GPT-5.4 nano scores higher on coding benchmarks: 43.1 versus 26.1 in the Noometry coding category.
How many benchmarks do GPT-5.4 nano and Phi 3 Mini 4k Instruct share?
18 benchmarks have published results for both models. GPT-5.4 nano has 40 scored results on Noometry and Phi 3 Mini 4k Instruct has 35.