Model comparison
gpt-oss-20b vs Qwen3 Max
Qwen3 Max is the stronger model overall, scoring 43.7 to 32.5 on the Noometry Index. gpt-oss-20b costs 67× less per token, which makes it the better buy when Qwen3 Max's lead doesn't matter for your workload.
Last verified . 24 shared benchmarks.
Alibaba (Qwen)Summary
- They share 24 benchmarks with published results for both. gpt-oss-20b scores higher in 1 category and Qwen3 Max in 7 categories; 7 gaps are clear of the uncertainty.
- The widest gap is in writing & preference, where Qwen3 Max leads 62.4 to 35.5.
- The biggest single-benchmark swing is Kagi LLM Benchmark: 53.2% for gpt-oss-20b and 72.5% for Qwen3 Max.
- gpt-oss-20b is cheaper at $0.018 / $0.09 per million input/output tokens, against $1.20 / $6 for Qwen3 Max.
- Qwen3 Max accepts more context: 262K tokens versus 131K.
- gpt-oss-20b has downloadable open weights; the other is API-only.
Side by side
| gpt-oss-20b | Qwen3 Max | |
|---|---|---|
| Provider | OpenAI | Alibaba (Qwen) |
| Noometry Index | 32.5 | 43.7 |
| Released | 2025-08-05 | 2025-09-23 |
| Weights | Open | Proprietary |
| Context window | 131K | 262K |
| Max output | 16K | 66K |
| Input $ / M tokens | $0.018 | $1.20 |
| Output $ / M tokens | $0.09 | $6 |
| Results tracked | 34 | 33 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding Qwen3 Max leads
gpt-oss-20b: 37.6 (#192), Qwen3 Max: 43.0 (#93)
| Benchmark | gpt-oss-20b | Qwen3 Max |
|---|---|---|
| LMArena Coding | 1306 | 1456 |
| ALE-Bench | 566.05 | 370.45 |
| SciCode | 34.4% | — |
| WeirdML | 40.9% | — |
Agentic & Tool Use Not comparable
gpt-oss-20b: 9.3 (#154), Qwen3 Max: —
| Benchmark | gpt-oss-20b | Qwen3 Max |
|---|---|---|
| Terminal-Bench | 3.4% | — |
| Vending-Bench 2 | — | 71.56 |
Reasoning Qwen3 Max leads
gpt-oss-20b: 19.3 (#261), Qwen3 Max: 22.6 (#190)
| Benchmark | gpt-oss-20b | Qwen3 Max |
|---|---|---|
| Kagi LLM Benchmark | 53.2% | 72.5% |
| Chess Puzzles | 4% | 4% |
| LMArena Hard Prompts | 1274 | 1448 |
| DTBench | 68% | 82.1% |
| LMCA | 14.5% | 28.3% |
| Epoch Capabilities Index | 137.82 | 142.38 |
| NYT Connections (extended) | — | 30.1% |
| CritPt | 1.4% | — |
| Mystery Game Puzzles | — | 5% |
Math Too close to call
gpt-oss-20b: 39.4 (#103), Qwen3 Max: 38.7 (#131)
| Benchmark | gpt-oss-20b | Qwen3 Max |
|---|---|---|
| OTIS Mock AIME 2024-2025 | 65.3% | 73.3% |
| LMArena Math | 1317 | 1446 |
| FrontierMath (Tiers 1-3) | — | 18.9% |
| Omni-MATH | 56.5% | — |
| MATH Level 5 | — | 97.1% |
Knowledge Qwen3 Max leads
gpt-oss-20b: 34.6 (#195), Qwen3 Max: 48.1 (#78)
| Benchmark | gpt-oss-20b | Qwen3 Max |
|---|---|---|
| GPQA Diamond | 60.8% | 72.6% |
| LMArena Expert | 1258 | 1455 |
| SimpleQA Verified | — | 48.7% |
| MMLU-Pro | 74% | — |
| GPQA (HELM) | 59.4% | — |
Multilingual Qwen3 Max leads
gpt-oss-20b: 42.2 (#197), Qwen3 Max: 53.7 (#62)
| Benchmark | gpt-oss-20b | Qwen3 Max |
|---|---|---|
| LMArena Non-English | 1268 | 1429 |
| LMArena Chinese | 1314 | 1478 |
| LMArena German | 1255 | 1463 |
| LMArena Japanese | 1244 | 1397 |
| LMArena Korean | 1236 | 1399 |
| LMArena Russian | 1278 | 1428 |
| LMArena Spanish | 1267 | 1462 |
| LMArena French | — | 1449 |
Instruction Following Qwen3 Max leads
gpt-oss-20b: 61.8 (#240), Qwen3 Max: 74.8 (#87)
| Benchmark | gpt-oss-20b | Qwen3 Max |
|---|---|---|
| LMArena Instruction Following | 1236 | 1419 |
| IFEval | 73.2% | — |
Long Context Qwen3 Max leads
gpt-oss-20b: 37.9 (#209), Qwen3 Max: 41.6 (#134)
| Benchmark | gpt-oss-20b | Qwen3 Max |
|---|---|---|
| LMArena Longer Query | 1250 | 1438 |
| Fiction.LiveBench | — | 66.7% |
| CL-bench | — | 14.5% |
Writing & Preference Qwen3 Max leads
gpt-oss-20b: 35.5 (#265), Qwen3 Max: 62.4 (#76)
| Benchmark | gpt-oss-20b | Qwen3 Max |
|---|---|---|
| LMArena Text | 1287 | 1439 |
| LMArena Creative Writing | 1201 | 1402 |
| LMArena Multi-Turn | 1268 | 1446 |
| EQ-Bench Creative Writing | 666 | — |
| WildBench | 73.7% | — |
Frequently asked questions
Is gpt-oss-20b better than Qwen3 Max?
Qwen3 Max is the stronger model overall, scoring 43.7 to 32.5 on the Noometry Index. gpt-oss-20b costs 67× less per token, which makes it the better buy when Qwen3 Max's lead doesn't matter for your workload.
Which is cheaper, gpt-oss-20b or Qwen3 Max?
gpt-oss-20b is cheaper. It lists at $0.018 per million input tokens and $0.09 per million output tokens; Qwen3 Max lists at $1.20 and $6.
Is gpt-oss-20b or Qwen3 Max better for coding?
Qwen3 Max scores higher on coding benchmarks: 43.0 versus 37.6 in the Noometry coding category.
Which has the bigger context window?
Qwen3 Max does, with 262K tokens against 131K.
How many benchmarks do gpt-oss-20b and Qwen3 Max share?
24 benchmarks have published results for both models. gpt-oss-20b has 34 scored results on Noometry and Qwen3 Max has 33.