Model comparison
o1 vs Qwen3.8 27B
Qwen3.8 27B is the stronger model overall, scoring 46.0 to 40.9 on the Noometry Index.
Last verified . 22 shared benchmarks.
Alibaba (Qwen)Summary
- They share 22 benchmarks with published results for both. o1 scores higher in 1 category and Qwen3.8 27B in 9 categories; 7 gaps are clear of the uncertainty.
- The widest gap is in reasoning, where Qwen3.8 27B leads 41.0 to 27.9.
- The biggest single-benchmark swing is ARC-AGI-1: 30.7% for o1 and 87.5% for Qwen3.8 27B.
- Qwen3.8 27B is cheaper at $0.04 / $2.30 per million input/output tokens, against $15 / $60 for o1.
- Qwen3.8 27B accepts more context: 262K tokens versus 200K.
- Qwen3.8 27B has downloadable open weights; the other is API-only.
Side by side
| o1 | Qwen3.8 27B | |
|---|---|---|
| Provider | OpenAI | Alibaba (Qwen) |
| Noometry Index | 40.9 | 46.0 |
| Released | 2024-09-12 | 2026-08-14 |
| Weights | Proprietary | Open |
| Context window | 200K | 262K |
| Max output | 100K | 33K |
| Input $ / M tokens | $15 | $0.04 |
| Output $ / M tokens | $60 | $2.30 |
| Results tracked | 52 | 31 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding Qwen3.8 27B leads
o1: 46.1 (#70), Qwen3.8 27B: 50.5 (#44)
| Benchmark | o1 | Qwen3.8 27B |
|---|---|---|
| LMArena Coding | 1367 | 1482 |
| Aider Polyglot | 61.7% | — |
| LMArena WebDev | — | 1593 |
| SciCode | — | 46.6% |
| WeirdML | 47.6% | — |
| LiveBench Coding | 69.7% | — |
| CadEval | 56% | — |
| HumanEval+ | 89% | — |
| MBPP+ | 80.2% | — |
Agentic & Tool Use Qwen3.8 27B leads
o1: 24.6 (#117), Qwen3.8 27B: 32.9 (#57)
| Benchmark | o1 | Qwen3.8 27B |
|---|---|---|
| APEX-Agents | — | 47.5% |
| Cybench | 10% | — |
| METR Time Horizons | 51.1% | — |
Reasoning Qwen3.8 27B leads
o1: 27.9 (#111), Qwen3.8 27B: 41.0 (#54)
| Benchmark | o1 | Qwen3.8 27B |
|---|---|---|
| ARC-AGI-1 | 30.7% | 87.5% |
| LMArena Hard Prompts | 1371 | 1460 |
| DTBench | 74.7% | 88% |
| LMCA | 22.3% | 41.4% |
| Epoch Capabilities Index | 141.91 | 149.38 |
| ARC-AGI-2 | — | 42.4% |
| SimpleBench | 41.7% | — |
| NYT Connections (extended) | — | 54.5% |
| CritPt | — | 5.4% |
| Chess Puzzles | 15% | — |
| EnigmaEval | 5.7% | — |
| LiveBench Reasoning | 91.6% | — |
| LiveBench Data Analysis | 65.5% | — |
| Surface Evolver Bench | — | 45% |
| LiveBench | 75.7% | — |
Math Too close to call
o1: 36.1 (#175), Qwen3.8 27B: 37.1 (#161)
| Benchmark | o1 | Qwen3.8 27B |
|---|---|---|
| LMArena Math | 1388 | 1456 |
| FrontierMath (Tiers 1-3) | 14.7% | — |
| OTIS Mock AIME 2024-2025 | 73.3% | — |
| ProofBench | — | 16% |
| LiveBench Math | 80.3% | — |
| MATH Level 5 | 94.7% | — |
| FrontierMath (Feb 2025 set) | 9.3% | — |
Knowledge Too close to call
o1: 41.5 (#110), Qwen3.8 27B: 41.6 (#109)
| Benchmark | o1 | Qwen3.8 27B |
|---|---|---|
| LMArena Expert | 1361 | 1482 |
| GPQA Diamond | 76.8% | — |
| Humanity's Last Exam | 8% | — |
| SimpleQA Verified | 41.1% | — |
| Confabulations | 11.7% | — |
Multimodal Qwen3.8 27B leads
o1: 34.2 (#93), Qwen3.8 27B: 41.3 (#37)
| Benchmark | o1 | Qwen3.8 27B |
|---|---|---|
| LMArena Vision | 1168 | 1271 |
| GeoBench | 80% | — |
| VPCT | 37% | — |
| SpatialViz-Bench | 41.4% | — |
Multilingual Qwen3.8 27B leads
o1: 48.6 (#142), Qwen3.8 27B: 53.7 (#60)
| Benchmark | o1 | Qwen3.8 27B |
|---|---|---|
| LMArena Non-English | 1358 | 1430 |
| LMArena Chinese | 1394 | 1504 |
| LMArena French | 1344 | 1465 |
| LMArena German | 1337 | 1438 |
| LMArena Japanese | 1346 | 1384 |
| LMArena Korean | 1396 | 1393 |
| LMArena Russian | 1356 | 1415 |
| LMArena Spanish | 1345 | 1448 |
Instruction Following Too close to call
o1: 74.8 (#86), Qwen3.8 27B: 75.8 (#53)
| Benchmark | o1 | Qwen3.8 27B |
|---|---|---|
| LMArena Instruction Following | 1367 | 1439 |
| LiveBench Instruction Following | 81.5% | — |
Long Context o1 leads
o1: 50.3 (#9), Qwen3.8 27B: 44.3 (#70)
| Benchmark | o1 | Qwen3.8 27B |
|---|---|---|
| LMArena Longer Query | 1378 | 1450 |
| Fiction.LiveBench | 83.3% | — |
Writing & Preference Qwen3.8 27B leads
o1: 55.6 (#144), Qwen3.8 27B: 65.8 (#43)
| Benchmark | o1 | Qwen3.8 27B |
|---|---|---|
| LMArena Text | 1366 | 1441 |
| LMArena Creative Writing | 1348 | 1384 |
| LMArena Multi-Turn | 1369 | 1441 |
| Short-Story Creative Writing | 70.2% | — |
| EQ-Bench Creative Writing | — | 1671 |
| LiveBench Language | 65.4% | — |
Frequently asked questions
Is o1 better than Qwen3.8 27B?
Qwen3.8 27B is the stronger model overall, scoring 46.0 to 40.9 on the Noometry Index.
Which is cheaper, o1 or Qwen3.8 27B?
Qwen3.8 27B is cheaper. It lists at $0.04 per million input tokens and $2.30 per million output tokens; o1 lists at $15 and $60.
Is o1 or Qwen3.8 27B better for coding?
Qwen3.8 27B scores higher on coding benchmarks: 50.5 versus 46.1 in the Noometry coding category.
Which has the bigger context window?
Qwen3.8 27B does, with 262K tokens against 200K.
How many benchmarks do o1 and Qwen3.8 27B share?
22 benchmarks have published results for both models. o1 has 52 scored results on Noometry and Qwen3.8 27B has 31.