Model comparison
GPT-5.4 mini vs o1
GPT-5.4 mini is the stronger model overall, scoring 45.0 to 40.9 on the Noometry Index.
Last verified . 29 shared benchmarks.
Summary
- They share 29 benchmarks with published results for both. GPT-5.4 mini scores higher in 7 categories and o1 in 3 categories; 8 gaps are clear of the uncertainty.
- The widest gap is in knowledge, where GPT-5.4 mini leads 51.5 to 41.5.
- The biggest single-benchmark swing is FrontierMath (Tiers 1-3): 51.2% for GPT-5.4 mini and 14.7% for o1.
- GPT-5.4 mini is cheaper at $0.75 / $4.50 per million input/output tokens, against $15 / $60 for o1.
- GPT-5.4 mini accepts more context: 400K tokens versus 200K.
Side by side
| GPT-5.4 mini | o1 | |
|---|---|---|
| Provider | OpenAI | OpenAI |
| Noometry Index | 45.0 | 40.9 |
| Released | 2026-03-17 | 2024-09-12 |
| Weights | Proprietary | Proprietary |
| Context window | 400K | 200K |
| Max output | 128K | 100K |
| Input $ / M tokens | $0.75 | $15 |
| Output $ / M tokens | $4.50 | $60 |
| Results tracked | 46 | 52 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding Too close to call
GPT-5.4 mini: 45.2 (#72), o1: 46.1 (#70)
| Benchmark | GPT-5.4 mini | o1 |
|---|---|---|
| WeirdML | 60.3% | 47.6% |
| LMArena Coding | 1438 | 1367 |
| FrontierCode | 27% | — |
| Aider Polyglot | — | 61.7% |
| LMArena WebDev | 1397 | — |
| SciCode | 49.9% | — |
| LiveBench Coding | — | 69.7% |
| CadEval | — | 56% |
| ALE-Bench | 1,189 | — |
| HumanEval+ | — | 89% |
| MBPP+ | — | 80.2% |
Agentic & Tool Use GPT-5.4 mini leads
GPT-5.4 mini: 29.9 (#81), o1: 24.6 (#117)
| Benchmark | GPT-5.4 mini | o1 |
|---|---|---|
| Cybench | — | 10% |
| DeepResearch Bench | 36.3% | — |
| METR Time Horizons | — | 51.1% |
Reasoning GPT-5.4 mini leads
GPT-5.4 mini: 30.4 (#85), o1: 27.9 (#111)
| Benchmark | GPT-5.4 mini | o1 |
|---|---|---|
| ARC-AGI-1 | 63.7% | 30.7% |
| Chess Puzzles | 24% | 15% |
| LMArena Hard Prompts | 1424 | 1371 |
| DTBench | 80% | 74.7% |
| LMCA | 40.8% | 22.3% |
| Epoch Capabilities Index | 148.84 | 141.91 |
| ARC-AGI-2 | 18.9% | — |
| SimpleBench | — | 41.7% |
| Kagi LLM Benchmark | 37.9% | — |
| NYT Connections (extended) | 61.8% | — |
| CritPt | 10% | — |
| EnigmaEval | — | 5.7% |
| Thematic Generalization | 61.7% | — |
| LiveBench Reasoning | — | 91.6% |
| Mystery Game Puzzles | 11% | — |
| LiveBench Data Analysis | — | 65.5% |
| ForecastBench | 57 | — |
| LiveBench | — | 75.7% |
Math GPT-5.4 mini leads
GPT-5.4 mini: 45.5 (#75), o1: 36.1 (#175)
| Benchmark | GPT-5.4 mini | o1 |
|---|---|---|
| FrontierMath (Tiers 1-3) | 51.2% | 14.7% |
| OTIS Mock AIME 2024-2025 | 88.9% | 73.3% |
| LMArena Math | 1419 | 1388 |
| FrontierMath (Feb 2025 set) | 28.3% | 9.3% |
| FrontierMath Tier 4 | 9.8% | — |
| ProofBench | 21% | — |
| LiveBench Math | — | 80.3% |
| MATH Level 5 | — | 94.7% |
| FrontierMath Tier 4 (v1) | 2.1% | — |
Knowledge GPT-5.4 mini leads
GPT-5.4 mini: 51.5 (#67), o1: 41.5 (#110)
| Benchmark | GPT-5.4 mini | o1 |
|---|---|---|
| GPQA Diamond | 86.9% | 76.8% |
| SimpleQA Verified | 29.4% | 41.1% |
| LMArena Expert | 1435 | 1361 |
| Humanity's Last Exam | — | 8% |
| Confabulations | — | 11.7% |
| Vectara Hallucination Rate | 5.5% | — |
Multimodal GPT-5.4 mini leads
GPT-5.4 mini: 39.7 (#56), o1: 34.2 (#93)
| Benchmark | GPT-5.4 mini | o1 |
|---|---|---|
| LMArena Vision | 1245 | 1168 |
| GeoBench | — | 80% |
| VPCT | — | 37% |
| SpatialViz-Bench | — | 41.4% |
Multilingual GPT-5.4 mini leads
GPT-5.4 mini: 51.9 (#96), o1: 48.6 (#142)
| Benchmark | GPT-5.4 mini | o1 |
|---|---|---|
| LMArena Non-English | 1405 | 1358 |
| LMArena Chinese | 1446 | 1394 |
| LMArena French | 1440 | 1344 |
| LMArena German | 1409 | 1337 |
| LMArena Japanese | 1374 | 1346 |
| LMArena Korean | 1368 | 1396 |
| LMArena Russian | 1417 | 1356 |
| LMArena Spanish | 1405 | 1345 |
Instruction Following Too close to call
GPT-5.4 mini: 74.1 (#102), o1: 74.8 (#86)
| Benchmark | GPT-5.4 mini | o1 |
|---|---|---|
| LMArena Instruction Following | 1405 | 1367 |
| LiveBench Instruction Following | — | 81.5% |
Long Context o1 leads
GPT-5.4 mini: 43.0 (#112), o1: 50.3 (#9)
| Benchmark | GPT-5.4 mini | o1 |
|---|---|---|
| LMArena Longer Query | 1407 | 1378 |
| Fiction.LiveBench | — | 83.3% |
Writing & Preference GPT-5.4 mini leads
GPT-5.4 mini: 64.0 (#58), o1: 55.6 (#144)
| Benchmark | GPT-5.4 mini | o1 |
|---|---|---|
| LMArena Text | 1412 | 1366 |
| LMArena Creative Writing | 1370 | 1348 |
| LMArena Multi-Turn | 1429 | 1369 |
| Short-Story Creative Writing | — | 70.2% |
| EQ-Bench Creative Writing | 1665 | — |
| LiveBench Language | — | 65.4% |
Frequently asked questions
Is GPT-5.4 mini better than o1?
GPT-5.4 mini is the stronger model overall, scoring 45.0 to 40.9 on the Noometry Index.
Which is cheaper, GPT-5.4 mini or o1?
GPT-5.4 mini is cheaper. It lists at $0.75 per million input tokens and $4.50 per million output tokens; o1 lists at $15 and $60.
Is GPT-5.4 mini or o1 better for coding?
They score almost the same on coding (45.2 vs 46.1); test both on your own repository before choosing.
Which has the bigger context window?
GPT-5.4 mini does, with 400K tokens against 200K.
How many benchmarks do GPT-5.4 mini and o1 share?
29 benchmarks have published results for both models. GPT-5.4 mini has 46 scored results on Noometry and o1 has 52.