Model comparison
GPT-5.4 vs o1
GPT-5.4 is the stronger model overall, scoring 59.4 to 40.9 on the Noometry Index.
Last verified . 32 shared benchmarks.
Summary
- They share 32 benchmarks with published results for both. GPT-5.4 scores higher in 9 categories and o1 in 1 category; 9 gaps are clear of the uncertainty.
- The widest gap is in math, where GPT-5.4 leads 73.5 to 36.1.
- The biggest single-benchmark swing is FrontierMath (Tiers 1-3): 78.6% for GPT-5.4 and 14.7% for o1.
- GPT-5.4 is cheaper at $2.50 / $15 per million input/output tokens, against $15 / $60 for o1.
- GPT-5.4 accepts more context: 1.05M tokens versus 200K.
Side by side
| GPT-5.4 | o1 | |
|---|---|---|
| Provider | OpenAI | OpenAI |
| Noometry Index | 59.4 | 40.9 |
| Released | 2026-03-05 | 2024-09-12 |
| Weights | Proprietary | Proprietary |
| Context window | 1.05M | 200K |
| Max output | 128K | 100K |
| Input $ / M tokens | $2.50 | $15 |
| Output $ / M tokens | $15 | $60 |
| Results tracked | 68 | 52 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding GPT-5.4 leads
GPT-5.4: 52.6 (#33), o1: 46.1 (#70)
| Benchmark | GPT-5.4 | o1 |
|---|---|---|
| WeirdML | 77.7% | 47.6% |
| LMArena Coding | 1497 | 1367 |
| SWE-bench Verified | 76.9% | — |
| DeepSWE | 51.8% | — |
| Aider Polyglot | — | 61.7% |
| LMArena WebDev | 1465 | — |
| SciCode | 56.6% | — |
| GSO | 31.4% | — |
| LiveBench Coding | — | 69.7% |
| MirrorCode | 15.6% | — |
| CadEval | — | 56% |
| ALE-Bench | 1,607 | — |
| AlgoTune | 1.85 | — |
| HumanEval+ | — | 89% |
| MBPP+ | — | 80.2% |
Agentic & Tool Use GPT-5.4 leads
GPT-5.4: 46.5 (#13), o1: 24.6 (#117)
| Benchmark | GPT-5.4 | o1 |
|---|---|---|
| METR Time Horizons | 74.3% | 51.1% |
| Terminal-Bench | 81.8% | — |
| APEX-Agents | 52.4% | — |
| τ²-bench Banking | 39.4% | — |
| Cybench | — | 10% |
| DeepResearch Bench | 35.1% | — |
| PostTrainBench | 19% | — |
| GBAEval | 45.1% | — |
| LMArena Search | 1197 | — |
| Vending-Bench 2 | 6,144 | — |
Reasoning GPT-5.4 leads
GPT-5.4: 61.8 (#19), o1: 27.9 (#111)
| Benchmark | GPT-5.4 | o1 |
|---|---|---|
| ARC-AGI-1 | 93.7% | 30.7% |
| Chess Puzzles | 44% | 15% |
| EnigmaEval | 16% | 5.7% |
| LMArena Hard Prompts | 1485 | 1371 |
| DTBench | 94.4% | 74.7% |
| LMCA | 52% | 22.3% |
| Epoch Capabilities Index | 156.81 | 141.91 |
| ARC-AGI-2 | 74% | — |
| SimpleBench | — | 41.7% |
| Kagi LLM Benchmark | 63.8% | — |
| NYT Connections (extended) | 91.3% | — |
| CritPt | 23.4% | — |
| Thematic Generalization | 80% | — |
| EBR-Bench | 25.4% | — |
| LiveBench Reasoning | — | 91.6% |
| Mystery Game Puzzles | 37% | — |
| LiveBench Data Analysis | — | 65.5% |
| ForecastBench | 59.5 | — |
| LiveBench | — | 75.7% |
Math GPT-5.4 leads
GPT-5.4: 73.5 (#19), o1: 36.1 (#175)
| Benchmark | GPT-5.4 | o1 |
|---|---|---|
| FrontierMath (Tiers 1-3) | 78.6% | 14.7% |
| OTIS Mock AIME 2024-2025 | 97.8% | 73.3% |
| LMArena Math | 1488 | 1388 |
| FrontierMath (Feb 2025 set) | 47.6% | 9.3% |
| FrontierMath Tier 4 | 49% | — |
| MathArena Final-Answer Competitions | 83.1% | — |
| ProofBench | 56% | — |
| LiveBench Math | — | 80.3% |
| MATH Level 5 | — | 94.7% |
| FrontierMath Tier 4 (v1) | 27.1% | — |
Knowledge GPT-5.4 leads
GPT-5.4: 65.3 (#14), o1: 41.5 (#110)
| Benchmark | GPT-5.4 | o1 |
|---|---|---|
| GPQA Diamond | 93.3% | 76.8% |
| Humanity's Last Exam | 36.2% | 8% |
| SimpleQA Verified | 45.1% | 41.1% |
| LMArena Expert | 1507 | 1361 |
| Confabulations | — | 11.7% |
| Vectara Hallucination Rate | 7% | — |
Multimodal GPT-5.4 leads
GPT-5.4: 43.7 (#20), o1: 34.2 (#93)
| Benchmark | GPT-5.4 | o1 |
|---|---|---|
| LMArena Vision | 1303 | 1168 |
| GeoBench | — | 80% |
| VPCT | — | 37% |
| Blueprint-Bench 2 | 27.1% | — |
| Furniture Assembly | 37.5% | — |
| LMArena Document | 1471 | — |
| SpatialViz-Bench | — | 41.4% |
Multilingual GPT-5.4 leads
GPT-5.4: 56.2 (#23), o1: 48.6 (#142)
| Benchmark | GPT-5.4 | o1 |
|---|---|---|
| LMArena Non-English | 1465 | 1358 |
| LMArena Chinese | 1519 | 1394 |
| LMArena French | 1493 | 1344 |
| LMArena German | 1472 | 1337 |
| LMArena Japanese | 1485 | 1346 |
| LMArena Korean | 1448 | 1396 |
| LMArena Russian | 1480 | 1356 |
| LMArena Spanish | 1454 | 1345 |
Instruction Following GPT-5.4 leads
GPT-5.4: 77.1 (#27), o1: 74.8 (#86)
| Benchmark | GPT-5.4 | o1 |
|---|---|---|
| LMArena Instruction Following | 1469 | 1367 |
| LiveBench Instruction Following | — | 81.5% |
Long Context Too close to call
GPT-5.4: 50.3 (#8), o1: 50.3 (#9)
| Benchmark | GPT-5.4 | o1 |
|---|---|---|
| LMArena Longer Query | 1473 | 1378 |
| Fiction.LiveBench | — | 83.3% |
| CL-bench | 27.9% | — |
| CL-bench Life | 21.7% | — |
Writing & Preference GPT-5.4 leads
GPT-5.4: 71.9 (#17), o1: 55.6 (#144)
| Benchmark | GPT-5.4 | o1 |
|---|---|---|
| LMArena Text | 1469 | 1366 |
| LMArena Creative Writing | 1439 | 1348 |
| LMArena Multi-Turn | 1482 | 1369 |
| Short-Story Creative Writing | — | 70.2% |
| EQ-Bench Creative Writing | 1840 | — |
| EQ-Bench 4 | 1272 | — |
| LiveBench Language | — | 65.4% |
Frequently asked questions
Is GPT-5.4 better than o1?
GPT-5.4 is the stronger model overall, scoring 59.4 to 40.9 on the Noometry Index.
Which is cheaper, GPT-5.4 or o1?
GPT-5.4 is cheaper. It lists at $2.50 per million input tokens and $15 per million output tokens; o1 lists at $15 and $60.
Is GPT-5.4 or o1 better for coding?
GPT-5.4 scores higher on coding benchmarks: 52.6 versus 46.1 in the Noometry coding category.
Which has the bigger context window?
GPT-5.4 does, with 1.05M tokens against 200K.
How many benchmarks do GPT-5.4 and o1 share?
32 benchmarks have published results for both models. GPT-5.4 has 68 scored results on Noometry and o1 has 52.