Model comparison
o1 vs Qwen2.5 72B Instruct
o1 is the stronger model overall, scoring 40.9 to 31.9 on the Noometry Index. Qwen2.5 72B Instruct costs 11× less per token, which makes it the better buy when o1's lead doesn't matter for your workload.
Last verified . 26 shared benchmarks.
Alibaba (Qwen)Summary
- They share 26 benchmarks with published results for both. o1 scores higher in 9 categories and Qwen2.5 72B Instruct in 0 categories; 9 gaps are clear of the uncertainty.
- The widest gap is in math, where o1 leads 36.1 to 19.3.
- The biggest single-benchmark swing is OTIS Mock AIME 2024-2025: 73.3% for o1 and 8.1% for Qwen2.5 72B Instruct.
- Qwen2.5 72B Instruct is cheaper at $1.40 / $5.60 per million input/output tokens, against $15 / $60 for o1.
- o1 accepts more context: 200K tokens versus 131K.
- Qwen2.5 72B Instruct has downloadable open weights; the other is API-only.
Side by side
| o1 | Qwen2.5 72B Instruct | |
|---|---|---|
| Provider | OpenAI | Alibaba (Qwen) |
| Noometry Index | 40.9 | 31.9 |
| Released | 2024-09-12 | 2024-09 |
| Weights | Proprietary | Open |
| Context window | 200K | 131K |
| Max output | 100K | 8K |
| Input $ / M tokens | $15 | $1.40 |
| Output $ / M tokens | $60 | $5.60 |
| Results tracked | 52 | 43 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding o1 leads
o1: 46.1 (#70), Qwen2.5 72B Instruct: 33.2 (#260)
| Benchmark | o1 | Qwen2.5 72B Instruct |
|---|---|---|
| WeirdML | 47.6% | 16% |
| LMArena Coding | 1367 | 1292 |
| Aider Polyglot | 61.7% | — |
| BigCodeBench Instruct | — | 45.8% |
| LiveBench Coding | 69.7% | — |
| BigCodeBench Complete | — | 55.9% |
| CadEval | 56% | — |
| HumanEval+ | 89% | — |
| MBPP+ | 80.2% | — |
Agentic & Tool Use o1 leads
o1: 24.6 (#117), Qwen2.5 72B Instruct: 22.1 (#133)
| Benchmark | o1 | Qwen2.5 72B Instruct |
|---|---|---|
| METR Time Horizons | 51.1% | 35.8% |
| TheAgentCompany | — | 5.7% |
| Cybench | 10% | — |
| BALROG | — | 16.2% |
Reasoning o1 leads
o1: 27.9 (#111), Qwen2.5 72B Instruct: 22.3 (#199)
| Benchmark | o1 | Qwen2.5 72B Instruct |
|---|---|---|
| LMArena Hard Prompts | 1371 | 1271 |
| DTBench | 74.7% | 62.9% |
| LMCA | 22.3% | 13.4% |
| Epoch Capabilities Index | 141.91 | 129 |
| SimpleBench | 41.7% | — |
| ARC-AGI-1 | 30.7% | — |
| Chess Puzzles | 15% | — |
| EnigmaEval | 5.7% | — |
| LiveBench Reasoning | 91.6% | — |
| LiveBench Data Analysis | 65.5% | — |
| BIG-Bench Hard | — | 79.8% |
| ForecastBench | — | 57.5 |
| HellaSwag | — | 84.8% |
| LiveBench | 75.7% | — |
| PIQA | — | 82.6% |
| WinoGrande | — | 82.3% |
Math o1 leads
o1: 36.1 (#175), Qwen2.5 72B Instruct: 19.3 (#287)
| Benchmark | o1 | Qwen2.5 72B Instruct |
|---|---|---|
| OTIS Mock AIME 2024-2025 | 73.3% | 8.1% |
| LMArena Math | 1388 | 1283 |
| MATH Level 5 | 94.7% | 63.2% |
| FrontierMath (Tiers 1-3) | 14.7% | — |
| Omni-MATH | — | 33% |
| LiveBench Math | 80.3% | — |
| FrontierMath (Feb 2025 set) | 9.3% | — |
Knowledge o1 leads
o1: 41.5 (#110), Qwen2.5 72B Instruct: 27.0 (#253)
| Benchmark | o1 | Qwen2.5 72B Instruct |
|---|---|---|
| GPQA Diamond | 76.8% | 49.1% |
| Confabulations | 11.7% | 19.1% |
| LMArena Expert | 1361 | 1245 |
| Humanity's Last Exam | 8% | — |
| SimpleQA Verified | 41.1% | — |
| MMLU-Pro | — | 63.1% |
| GPQA (HELM) | — | 42.6% |
| ARC (AI2) Challenge | — | 94.5% |
| MMLU | — | 85.3% |
| TriviaQA | — | 71.9% |
Multimodal Not comparable
o1: 34.2 (#93), Qwen2.5 72B Instruct: —
| Benchmark | o1 | Qwen2.5 72B Instruct |
|---|---|---|
| LMArena Vision | 1168 | — |
| GeoBench | 80% | — |
| VPCT | 37% | — |
| SpatialViz-Bench | 41.4% | — |
Multilingual o1 leads
o1: 48.6 (#142), Qwen2.5 72B Instruct: 41.0 (#213)
| Benchmark | o1 | Qwen2.5 72B Instruct |
|---|---|---|
| LMArena Non-English | 1358 | 1252 |
| LMArena Chinese | 1394 | 1272 |
| LMArena French | 1344 | 1280 |
| LMArena German | 1337 | 1234 |
| LMArena Japanese | 1346 | 1180 |
| LMArena Korean | 1396 | 1188 |
| LMArena Russian | 1356 | 1264 |
| LMArena Spanish | 1345 | 1256 |
Instruction Following o1 leads
o1: 74.8 (#86), Qwen2.5 72B Instruct: 65.5 (#221)
| Benchmark | o1 | Qwen2.5 72B Instruct |
|---|---|---|
| LMArena Instruction Following | 1367 | 1254 |
| LiveBench Instruction Following | 81.5% | — |
| IFEval | — | 80.6% |
Long Context o1 leads
o1: 50.3 (#9), Qwen2.5 72B Instruct: 38.9 (#188)
| Benchmark | o1 | Qwen2.5 72B Instruct |
|---|---|---|
| LMArena Longer Query | 1378 | 1282 |
| Fiction.LiveBench | 83.3% | — |
Writing & Preference o1 leads
o1: 55.6 (#144), Qwen2.5 72B Instruct: 46.7 (#215)
| Benchmark | o1 | Qwen2.5 72B Instruct |
|---|---|---|
| LMArena Text | 1366 | 1269 |
| LMArena Creative Writing | 1348 | 1221 |
| LMArena Multi-Turn | 1369 | 1272 |
| Short-Story Creative Writing | 70.2% | — |
| WildBench | — | 80.2% |
| LiveBench Language | 65.4% | — |
Frequently asked questions
Is o1 better than Qwen2.5 72B Instruct?
o1 is the stronger model overall, scoring 40.9 to 31.9 on the Noometry Index. Qwen2.5 72B Instruct costs 11× less per token, which makes it the better buy when o1's lead doesn't matter for your workload.
Which is cheaper, o1 or Qwen2.5 72B Instruct?
Qwen2.5 72B Instruct is cheaper. It lists at $1.40 per million input tokens and $5.60 per million output tokens; o1 lists at $15 and $60.
Is o1 or Qwen2.5 72B Instruct better for coding?
o1 scores higher on coding benchmarks: 46.1 versus 33.2 in the Noometry coding category.
Which has the bigger context window?
o1 does, with 200K tokens against 131K.
How many benchmarks do o1 and Qwen2.5 72B Instruct share?
26 benchmarks have published results for both models. o1 has 52 scored results on Noometry and Qwen2.5 72B Instruct has 43.