Model comparison
gpt-oss-20b vs Qwen3.5 35B-A3B
Qwen3.5 35B-A3B is the stronger model overall, scoring 42.0 to 32.5 on the Noometry Index. gpt-oss-20b costs 19× less per token, which makes it the better buy when Qwen3.5 35B-A3B'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 Qwen3.5 35B-A3B in 7 categories; 7 gaps are clear of the uncertainty.
- The widest gap is in writing & preference, where Qwen3.5 35B-A3B leads 57.9 to 35.5.
- The biggest single-benchmark swing is GPQA Diamond: 60.8% for gpt-oss-20b and 83.5% for Qwen3.5 35B-A3B.
- gpt-oss-20b is cheaper at $0.018 / $0.09 per million input/output tokens, against $0.25 / $2 for Qwen3.5 35B-A3B.
- Qwen3.5 35B-A3B accepts more context: 262K tokens versus 131K.
Side by side
| gpt-oss-20b | Qwen3.5 35B-A3B | |
|---|---|---|
| Provider | OpenAI | Alibaba (Qwen) |
| Noometry Index | 32.5 | 42.0 |
| Released | 2025-08-05 | 2026-02-01 |
| Weights | Open | Open |
| Context window | 131K | 262K |
| Max output | 16K | 66K |
| Input $ / M tokens | $0.018 | $0.25 |
| Output $ / M tokens | $0.09 | $2 |
| Results tracked | 34 | 28 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding gpt-oss-20b leads
gpt-oss-20b: 37.6 (#192), Qwen3.5 35B-A3B: 33.8 (#251)
| Benchmark | gpt-oss-20b | Qwen3.5 35B-A3B |
|---|---|---|
| SciCode | 34.4% | 29.3% |
| LMArena Coding | 1306 | 1410 |
| LMArena WebDev | — | 1254 |
| WeirdML | 40.9% | — |
| ALE-Bench | 566.05 | — |
Agentic & Tool Use Not comparable
gpt-oss-20b: 9.3 (#154), Qwen3.5 35B-A3B: —
| Benchmark | gpt-oss-20b | Qwen3.5 35B-A3B |
|---|---|---|
| Terminal-Bench | 3.4% | — |
Reasoning Qwen3.5 35B-A3B leads
gpt-oss-20b: 19.3 (#261), Qwen3.5 35B-A3B: 24.6 (#161)
| Benchmark | gpt-oss-20b | Qwen3.5 35B-A3B |
|---|---|---|
| CritPt | 1.4% | 0.6% |
| Chess Puzzles | 4% | 10% |
| LMArena Hard Prompts | 1274 | 1400 |
| DTBench | 68% | 80% |
| LMCA | 14.5% | 29.5% |
| Epoch Capabilities Index | 137.82 | 142.52 |
| Kagi LLM Benchmark | 53.2% | — |
Math Too close to call
gpt-oss-20b: 39.4 (#103), Qwen3.5 35B-A3B: 39.9 (#97)
| Benchmark | gpt-oss-20b | Qwen3.5 35B-A3B |
|---|---|---|
| OTIS Mock AIME 2024-2025 | 65.3% | 70% |
| LMArena Math | 1317 | 1404 |
| MathArena Final-Answer Competitions | — | 56% |
| Omni-MATH | 56.5% | — |
Knowledge Qwen3.5 35B-A3B leads
gpt-oss-20b: 34.6 (#195), Qwen3.5 35B-A3B: 47.8 (#79)
| Benchmark | gpt-oss-20b | Qwen3.5 35B-A3B |
|---|---|---|
| GPQA Diamond | 60.8% | 83.5% |
| LMArena Expert | 1258 | 1408 |
| MMLU-Pro | 74% | — |
| Vectara Hallucination Rate | — | 10.5% |
| GPQA (HELM) | 59.4% | — |
Multilingual Qwen3.5 35B-A3B leads
gpt-oss-20b: 42.2 (#197), Qwen3.5 35B-A3B: 50.0 (#127)
| Benchmark | gpt-oss-20b | Qwen3.5 35B-A3B |
|---|---|---|
| LMArena Non-English | 1268 | 1378 |
| LMArena Chinese | 1314 | 1457 |
| LMArena German | 1255 | 1367 |
| LMArena Japanese | 1244 | 1325 |
| LMArena Korean | 1236 | 1356 |
| LMArena Russian | 1278 | 1376 |
| LMArena Spanish | 1267 | 1392 |
| LMArena French | — | 1412 |
Instruction Following Qwen3.5 35B-A3B leads
gpt-oss-20b: 61.8 (#240), Qwen3.5 35B-A3B: 72.8 (#128)
| Benchmark | gpt-oss-20b | Qwen3.5 35B-A3B |
|---|---|---|
| LMArena Instruction Following | 1236 | 1379 |
| IFEval | 73.2% | — |
Long Context Qwen3.5 35B-A3B leads
gpt-oss-20b: 37.9 (#209), Qwen3.5 35B-A3B: 42.4 (#127)
| Benchmark | gpt-oss-20b | Qwen3.5 35B-A3B |
|---|---|---|
| LMArena Longer Query | 1250 | 1389 |
Writing & Preference Qwen3.5 35B-A3B leads
gpt-oss-20b: 35.5 (#265), Qwen3.5 35B-A3B: 57.9 (#124)
| Benchmark | gpt-oss-20b | Qwen3.5 35B-A3B |
|---|---|---|
| LMArena Text | 1287 | 1395 |
| LMArena Creative Writing | 1201 | 1346 |
| LMArena Multi-Turn | 1268 | 1390 |
| EQ-Bench Creative Writing | 666 | — |
| WildBench | 73.7% | — |
Frequently asked questions
Is gpt-oss-20b better than Qwen3.5 35B-A3B?
Qwen3.5 35B-A3B is the stronger model overall, scoring 42.0 to 32.5 on the Noometry Index. gpt-oss-20b costs 19× less per token, which makes it the better buy when Qwen3.5 35B-A3B's lead doesn't matter for your workload.
Which is cheaper, gpt-oss-20b or Qwen3.5 35B-A3B?
gpt-oss-20b is cheaper. It lists at $0.018 per million input tokens and $0.09 per million output tokens; Qwen3.5 35B-A3B lists at $0.25 and $2.
Is gpt-oss-20b or Qwen3.5 35B-A3B better for coding?
gpt-oss-20b scores higher on coding benchmarks: 37.6 versus 33.8 in the Noometry coding category.
Which has the bigger context window?
Qwen3.5 35B-A3B does, with 262K tokens against 131K.
How many benchmarks do gpt-oss-20b and Qwen3.5 35B-A3B share?
24 benchmarks have published results for both models. gpt-oss-20b has 34 scored results on Noometry and Qwen3.5 35B-A3B has 28.