Model comparison
GPT-5 Nano vs o1
o1 is the stronger model overall, scoring 40.9 to 33.5 on the Noometry Index. GPT-5 Nano costs 191× less per token, which makes it the better buy when o1's lead doesn't matter for your workload.
Last verified . 31 shared benchmarks.
Summary
- They share 31 benchmarks with published results for both. GPT-5 Nano scores higher in 2 categories and o1 in 8 categories; 9 gaps are clear of the uncertainty.
- The widest gap is in long context, where o1 leads 50.3 to 31.3.
- The biggest single-benchmark swing is Fiction.LiveBench: 44.4% for GPT-5 Nano and 83.3% for o1.
- GPT-5 Nano is cheaper at $0.05 / $0.40 per million input/output tokens, against $15 / $60 for o1.
- GPT-5 Nano accepts more context: 400K tokens versus 200K.
Side by side
| GPT-5 Nano | o1 | |
|---|---|---|
| Provider | OpenAI | OpenAI |
| Noometry Index | 33.5 | 40.9 |
| Released | 2025-08-07 | 2024-09-12 |
| Weights | Proprietary | Proprietary |
| Context window | 400K | 200K |
| Max output | 128K | 100K |
| Input $ / M tokens | $0.05 | $15 |
| Output $ / M tokens | $0.40 | $60 |
| Results tracked | 49 | 52 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding o1 leads
GPT-5 Nano: 33.6 (#254), o1: 46.1 (#70)
| Benchmark | GPT-5 Nano | o1 |
|---|---|---|
| WeirdML | 38.1% | 47.6% |
| LMArena Coding | 1351 | 1367 |
| SWE-bench Verified (bash only) | 34.8% | — |
| Aider Polyglot | — | 61.7% |
| LiveBench Coding | — | 69.7% |
| CadEval | — | 56% |
| ALE-Bench | 718.67 | — |
| HumanEval+ | — | 89% |
| MBPP+ | — | 80.2% |
Agentic & Tool Use GPT-5 Nano leads
GPT-5 Nano: 25.8 (#106), o1: 24.6 (#117)
| Benchmark | GPT-5 Nano | o1 |
|---|---|---|
| Terminal-Bench | 21.8% | — |
| Berkeley Function Calling Leaderboard | 51.5% | — |
| Cybench | — | 10% |
| METR Time Horizons | — | 51.1% |
Reasoning o1 leads
GPT-5 Nano: 16.3 (#306), o1: 27.9 (#111)
| Benchmark | GPT-5 Nano | o1 |
|---|---|---|
| ARC-AGI-1 | 20.7% | 30.7% |
| Chess Puzzles | 27% | 15% |
| LMArena Hard Prompts | 1328 | 1371 |
| DTBench | 62.7% | 74.7% |
| LMCA | 7.9% | 22.3% |
| Epoch Capabilities Index | 139.38 | 141.91 |
| ARC-AGI-2 | 2.6% | — |
| SimpleBench | — | 41.7% |
| Kagi LLM Benchmark | 62.2% | — |
| EnigmaEval | — | 5.7% |
| LiveBench Reasoning | — | 91.6% |
| Mystery Game Puzzles | 9% | — |
| LiveBench Data Analysis | — | 65.5% |
| ForecastBench | 59.1 | — |
| LiveBench | — | 75.7% |
Math o1 leads
GPT-5 Nano: 29.4 (#241), o1: 36.1 (#175)
| Benchmark | GPT-5 Nano | o1 |
|---|---|---|
| FrontierMath (Tiers 1-3) | 20% | 14.7% |
| OTIS Mock AIME 2024-2025 | 81.1% | 73.3% |
| LMArena Math | 1317 | 1388 |
| MATH Level 5 | 95.2% | 94.7% |
| FrontierMath (Feb 2025 set) | 8.3% | 9.3% |
| FrontierMath Tier 4 | 2.4% | — |
| ProofBench | 12% | — |
| Omni-MATH | 54.6% | — |
| LiveBench Math | — | 80.3% |
| FrontierMath Tier 4 (v1) | 2.1% | — |
Knowledge o1 leads
GPT-5 Nano: 35.9 (#178), o1: 41.5 (#110)
| Benchmark | GPT-5 Nano | o1 |
|---|---|---|
| GPQA Diamond | 69.4% | 76.8% |
| SimpleQA Verified | 11.7% | 41.1% |
| LMArena Expert | 1321 | 1361 |
| Humanity's Last Exam | — | 8% |
| MMLU-Pro | 77.8% | — |
| Confabulations | — | 11.7% |
| Vectara Hallucination Rate | 10.5% | — |
| GPQA (HELM) | 67.9% | — |
Multimodal o1 leads
GPT-5 Nano: 31.3 (#108), o1: 34.2 (#93)
| Benchmark | GPT-5 Nano | o1 |
|---|---|---|
| LMArena Vision | 1159 | 1168 |
| VPCT | 37.2% | 37% |
| GeoBench | — | 80% |
| SpatialViz-Bench | — | 41.4% |
Multilingual o1 leads
GPT-5 Nano: 45.3 (#172), o1: 48.6 (#142)
| Benchmark | GPT-5 Nano | o1 |
|---|---|---|
| LMArena Non-English | 1313 | 1358 |
| LMArena Chinese | 1356 | 1394 |
| LMArena German | 1327 | 1337 |
| LMArena Japanese | 1226 | 1346 |
| LMArena Korean | 1269 | 1396 |
| LMArena Russian | 1296 | 1356 |
| LMArena Spanish | 1360 | 1345 |
| LMArena French | — | 1344 |
Instruction Following Too close to call
GPT-5 Nano: 75.0 (#79), o1: 74.8 (#86)
| Benchmark | GPT-5 Nano | o1 |
|---|---|---|
| LMArena Instruction Following | 1306 | 1367 |
| LiveBench Instruction Following | — | 81.5% |
| IFEval | 93.2% | — |
Long Context o1 leads
GPT-5 Nano: 31.3 (#281), o1: 50.3 (#9)
| Benchmark | GPT-5 Nano | o1 |
|---|---|---|
| Fiction.LiveBench | 44.4% | 83.3% |
| LMArena Longer Query | 1312 | 1378 |
Writing & Preference o1 leads
GPT-5 Nano: 39.1 (#249), o1: 55.6 (#144)
| Benchmark | GPT-5 Nano | o1 |
|---|---|---|
| LMArena Text | 1320 | 1366 |
| LMArena Creative Writing | 1249 | 1348 |
| LMArena Multi-Turn | 1311 | 1369 |
| Short-Story Creative Writing | — | 70.2% |
| EQ-Bench Creative Writing | 705 | — |
| WildBench | 80.6% | — |
| LiveBench Language | — | 65.4% |
Frequently asked questions
Is GPT-5 Nano better than o1?
o1 is the stronger model overall, scoring 40.9 to 33.5 on the Noometry Index. GPT-5 Nano costs 191× less per token, which makes it the better buy when o1's lead doesn't matter for your workload.
Which is cheaper, GPT-5 Nano or o1?
GPT-5 Nano is cheaper. It lists at $0.05 per million input tokens and $0.40 per million output tokens; o1 lists at $15 and $60.
Is GPT-5 Nano or o1 better for coding?
o1 scores higher on coding benchmarks: 46.1 versus 33.6 in the Noometry coding category.
Which has the bigger context window?
GPT-5 Nano does, with 400K tokens against 200K.
How many benchmarks do GPT-5 Nano and o1 share?
31 benchmarks have published results for both models. GPT-5 Nano has 49 scored results on Noometry and o1 has 52.