Model comparison
o1 vs Qwen3 32B
o1 is the stronger model overall, scoring 40.9 to 39.2 on the Noometry Index. Qwen3 32B costs 21× less per token, which makes it the better buy when o1's lead doesn't matter for your workload.
Last verified . 21 shared benchmarks.
Alibaba (Qwen)Summary
- They share 21 benchmarks with published results for both. o1 scores higher in 7 categories and Qwen3 32B in 2 categories; 9 gaps are clear of the uncertainty.
- The widest gap is in coding, where o1 leads 46.1 to 37.7.
- The biggest single-benchmark swing is Aider Polyglot: 61.7% for o1 and 40% for Qwen3 32B.
- Qwen3 32B is cheaper at $0.70 / $2.80 per million input/output tokens, against $15 / $60 for o1.
- o1 accepts more context: 200K tokens versus 131K.
- Qwen3 32B has downloadable open weights; the other is API-only.
Side by side
| o1 | Qwen3 32B | |
|---|---|---|
| Provider | OpenAI | Alibaba (Qwen) |
| Noometry Index | 40.9 | 39.2 |
| Released | 2024-09-12 | 2025-04 |
| Weights | Proprietary | Open |
| Context window | 200K | 131K |
| Max output | 100K | 16K |
| Input $ / M tokens | $15 | $0.70 |
| Output $ / M tokens | $60 | $2.80 |
| Results tracked | 52 | 26 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding o1 leads
o1: 46.1 (#70), Qwen3 32B: 37.7 (#190)
| Benchmark | o1 | Qwen3 32B |
|---|---|---|
| Aider Polyglot | 61.7% | 40% |
| LMArena Coding | 1367 | 1358 |
| SciCode | — | 35.4% |
| WeirdML | 47.6% | — |
| LiveBench Coding | 69.7% | — |
| CadEval | 56% | — |
| HumanEval+ | 89% | — |
| MBPP+ | 80.2% | — |
Agentic & Tool Use Qwen3 32B leads
o1: 24.6 (#117), Qwen3 32B: 32.6 (#62)
| Benchmark | o1 | Qwen3 32B |
|---|---|---|
| Berkeley Function Calling Leaderboard | — | 48.7% |
| Cybench | 10% | — |
| METR Time Horizons | 51.1% | — |
Reasoning o1 leads
o1: 27.9 (#111), Qwen3 32B: 20.2 (#241)
| Benchmark | o1 | Qwen3 32B |
|---|---|---|
| Chess Puzzles | 15% | 5% |
| LMArena Hard Prompts | 1371 | 1334 |
| DTBench | 74.7% | 67.5% |
| LMCA | 22.3% | 17.3% |
| Epoch Capabilities Index | 141.91 | 138.51 |
| SimpleBench | 41.7% | — |
| Kagi LLM Benchmark | — | 54.9% |
| ARC-AGI-1 | 30.7% | — |
| CritPt | — | 0.3% |
| EnigmaEval | 5.7% | — |
| LiveBench Reasoning | 91.6% | — |
| LiveBench Data Analysis | 65.5% | — |
| LiveBench | 75.7% | — |
Math Qwen3 32B leads
o1: 36.1 (#175), Qwen3 32B: 39.7 (#99)
| Benchmark | o1 | Qwen3 32B |
|---|---|---|
| OTIS Mock AIME 2024-2025 | 73.3% | 66.9% |
| LMArena Math | 1388 | 1399 |
| FrontierMath (Tiers 1-3) | 14.7% | — |
| LiveBench Math | 80.3% | — |
| MATH Level 5 | 94.7% | — |
| FrontierMath (Feb 2025 set) | 9.3% | — |
Knowledge o1 leads
o1: 41.5 (#110), Qwen3 32B: 40.0 (#125)
| Benchmark | o1 | Qwen3 32B |
|---|---|---|
| GPQA Diamond | 76.8% | 65.7% |
| LMArena Expert | 1361 | 1362 |
| Humanity's Last Exam | 8% | — |
| SimpleQA Verified | 41.1% | — |
| Confabulations | 11.7% | — |
| Vectara Hallucination Rate | — | 5.9% |
Multimodal Not comparable
o1: 34.2 (#93), Qwen3 32B: —
| Benchmark | o1 | Qwen3 32B |
|---|---|---|
| LMArena Vision | 1168 | — |
| GeoBench | 80% | — |
| VPCT | 37% | — |
| SpatialViz-Bench | 41.4% | — |
Multilingual o1 leads
o1: 48.6 (#142), Qwen3 32B: 45.6 (#167)
| Benchmark | o1 | Qwen3 32B |
|---|---|---|
| LMArena Non-English | 1358 | 1317 |
| LMArena Chinese | 1394 | 1357 |
| LMArena German | 1337 | 1341 |
| LMArena Russian | 1356 | 1311 |
| LMArena French | 1344 | — |
| LMArena Japanese | 1346 | — |
| LMArena Korean | 1396 | — |
| LMArena Spanish | 1345 | — |
Instruction Following o1 leads
o1: 74.8 (#86), Qwen3 32B: 68.9 (#179)
| Benchmark | o1 | Qwen3 32B |
|---|---|---|
| LMArena Instruction Following | 1367 | 1305 |
| LiveBench Instruction Following | 81.5% | — |
Long Context o1 leads
o1: 50.3 (#9), Qwen3 32B: 43.8 (#87)
| Benchmark | o1 | Qwen3 32B |
|---|---|---|
| Fiction.LiveBench | 83.3% | 74.2% |
| LMArena Longer Query | 1378 | 1327 |
Writing & Preference o1 leads
o1: 55.6 (#144), Qwen3 32B: 52.9 (#163)
| Benchmark | o1 | Qwen3 32B |
|---|---|---|
| LMArena Text | 1366 | 1340 |
| LMArena Creative Writing | 1348 | 1297 |
| LMArena Multi-Turn | 1369 | 1331 |
| Short-Story Creative Writing | 70.2% | — |
| LiveBench Language | 65.4% | — |
Frequently asked questions
Is o1 better than Qwen3 32B?
o1 is the stronger model overall, scoring 40.9 to 39.2 on the Noometry Index. Qwen3 32B costs 21× less per token, which makes it the better buy when o1's lead doesn't matter for your workload.
Which is cheaper, o1 or Qwen3 32B?
Qwen3 32B is cheaper. It lists at $0.70 per million input tokens and $2.80 per million output tokens; o1 lists at $15 and $60.
Is o1 or Qwen3 32B better for coding?
o1 scores higher on coding benchmarks: 46.1 versus 37.7 in the Noometry coding category.
Which has the bigger context window?
o1 does, with 200K tokens against 131K.
How many benchmarks do o1 and Qwen3 32B share?
21 benchmarks have published results for both models. o1 has 52 scored results on Noometry and Qwen3 32B has 26.