Model comparison
GPT-5.4 nano vs o3-mini
GPT-5.4 nano is the stronger model overall, scoring 41.9 to 36.7 on the Noometry Index.
Last verified . 35 shared benchmarks.
Summary
- They share 35 benchmarks with published results for both. GPT-5.4 nano scores higher in 7 categories and o3-mini in 1 category; 8 gaps are clear of the uncertainty.
- The widest gap is in math, where GPT-5.4 nano leads 40.7 to 27.6.
- The biggest single-benchmark swing is FrontierMath (Tiers 1-3): 44.9% for GPT-5.4 nano and 18.6% for o3-mini.
- GPT-5.4 nano is cheaper at $0.20 / $1.25 per million input/output tokens, against $1.10 / $4.40 for o3-mini.
- GPT-5.4 nano accepts more context: 400K tokens versus 200K.
Side by side
| GPT-5.4 nano | o3-mini | |
|---|---|---|
| Provider | OpenAI | OpenAI |
| Noometry Index | 41.9 | 36.7 |
| Released | 2026-03-17 | 2024-12-20 |
| Weights | Proprietary | Proprietary |
| Context window | 400K | 200K |
| Max output | 128K | 100K |
| Input $ / M tokens | $0.20 | $1.10 |
| Output $ / M tokens | $1.25 | $4.40 |
| Results tracked | 40 | 51 |
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), o3-mini: 41.4 (#116)
| Benchmark | GPT-5.4 nano | o3-mini |
|---|---|---|
| SciCode | 46.9% | 39.8% |
| WeirdML | 49.2% | 43.7% |
| LMArena Coding | 1404 | 1378 |
| Aider Polyglot | — | 60.4% |
| GSO | — | 1.3% |
| LiveBench Coding | — | 82.7% |
| CadEval | — | 54% |
| ALE-Bench | 1,005 | — |
Agentic & Tool Use Not comparable
GPT-5.4 nano: —, o3-mini: 29.3 (#79)
| Benchmark | GPT-5.4 nano | o3-mini |
|---|---|---|
| Cybench | — | 22.5% |
Reasoning GPT-5.4 nano leads
GPT-5.4 nano: 23.8 (#159), o3-mini: 16.3 (#295)
| Benchmark | GPT-5.4 nano | o3-mini |
|---|---|---|
| ARC-AGI-2 | 5.7% | 3% |
| ARC-AGI-1 | 51.5% | 34.5% |
| CritPt | 9.3% | 0.3% |
| Chess Puzzles | 30% | 17% |
| LMArena Hard Prompts | 1380 | 1366 |
| Mystery Game Puzzles | 9% | 7% |
| DTBench | 80.3% | 68.8% |
| LMCA | 36.9% | 19% |
| Epoch Capabilities Index | 145.78 | 140.35 |
| ForecastBench | 57.3 | 59.6 |
| SimpleBench | — | 22.8% |
| Kagi LLM Benchmark | 39.7% | — |
| LiveBench Reasoning | — | 89.6% |
| LiveBench Data Analysis | — | 70.6% |
| LiveBench | — | 75.9% |
Math GPT-5.4 nano leads
GPT-5.4 nano: 40.7 (#76), o3-mini: 27.6 (#236)
| Benchmark | GPT-5.4 nano | o3-mini |
|---|---|---|
| FrontierMath (Tiers 1-3) | 44.9% | 18.6% |
| FrontierMath Tier 4 | 12.2% | 0% |
| OTIS Mock AIME 2024-2025 | 87.8% | 76.9% |
| LMArena Math | 1406 | 1396 |
| FrontierMath (Feb 2025 set) | 25.9% | 12.4% |
| FrontierMath Tier 4 (v1) | 6.3% | 4.2% |
| ProofBench | 5% | — |
| LiveBench Math | — | 77.3% |
| MATH Level 5 | — | 96.5% |
Knowledge GPT-5.4 nano leads
GPT-5.4 nano: 41.8 (#94), o3-mini: 38.3 (#135)
| Benchmark | GPT-5.4 nano | o3-mini |
|---|---|---|
| GPQA Diamond | 78.5% | 77% |
| SimpleQA Verified | 11.7% | 15.3% |
| LMArena Expert | 1397 | 1364 |
| Confabulations | — | 17.9% |
| Vectara Hallucination Rate | 3.1% | — |
Multimodal Not comparable
GPT-5.4 nano: 36.0 (#73), o3-mini: —
| Benchmark | GPT-5.4 nano | o3-mini |
|---|---|---|
| LMArena Vision | 1196 | — |
Multilingual GPT-5.4 nano leads
GPT-5.4 nano: 48.8 (#136), o3-mini: 45.9 (#159)
| Benchmark | GPT-5.4 nano | o3-mini |
|---|---|---|
| LMArena Non-English | 1359 | 1319 |
| LMArena Chinese | 1391 | 1379 |
| LMArena French | 1396 | 1333 |
| LMArena German | 1368 | 1303 |
| LMArena Japanese | 1342 | 1286 |
| LMArena Korean | 1321 | 1314 |
| LMArena Russian | 1363 | 1304 |
| LMArena Spanish | 1368 | 1321 |
Instruction Following o3-mini leads
GPT-5.4 nano: 72.1 (#139), o3-mini: 75.3 (#67)
| Benchmark | GPT-5.4 nano | o3-mini |
|---|---|---|
| LMArena Instruction Following | 1362 | 1338 |
| LiveBench Instruction Following | — | 84.4% |
Long Context GPT-5.4 nano leads
GPT-5.4 nano: 41.6 (#131), o3-mini: 33.8 (#251)
| Benchmark | GPT-5.4 nano | o3-mini |
|---|---|---|
| LMArena Longer Query | 1365 | 1343 |
| Fiction.LiveBench | — | 50% |
Writing & Preference GPT-5.4 nano leads
GPT-5.4 nano: 55.9 (#137), o3-mini: 50.4 (#177)
| Benchmark | GPT-5.4 nano | o3-mini |
|---|---|---|
| LMArena Text | 1372 | 1337 |
| LMArena Creative Writing | 1313 | 1286 |
| LMArena Multi-Turn | 1382 | 1320 |
| Short-Story Creative Writing | — | 61.7% |
| LiveBench Language | — | 50.7% |
Frequently asked questions
Is GPT-5.4 nano better than o3-mini?
GPT-5.4 nano is the stronger model overall, scoring 41.9 to 36.7 on the Noometry Index.
Which is cheaper, GPT-5.4 nano or o3-mini?
GPT-5.4 nano is cheaper. It lists at $0.20 per million input tokens and $1.25 per million output tokens; o3-mini lists at $1.10 and $4.40.
Is GPT-5.4 nano or o3-mini better for coding?
GPT-5.4 nano scores higher on coding benchmarks: 43.1 versus 41.4 in the Noometry coding category.
Which has the bigger context window?
GPT-5.4 nano does, with 400K tokens against 200K.
How many benchmarks do GPT-5.4 nano and o3-mini share?
35 benchmarks have published results for both models. GPT-5.4 nano has 40 scored results on Noometry and o3-mini has 51.