Model comparison
gpt-oss-120b vs Qwen2-72B
gpt-oss-120b is the stronger model overall, scoring 36.3 to 30.0 on the Noometry Index.
Last verified . 21 shared benchmarks.
Alibaba (Qwen)Summary
- They share 21 benchmarks with published results for both. gpt-oss-120b scores higher in 6 categories and Qwen2-72B in 3 categories; 9 gaps are clear of the uncertainty.
- The widest gap is in math, where gpt-oss-120b leads 52.5 to 30.2.
- The biggest single-benchmark swing is WeirdML: 48.2% for gpt-oss-120b and 11.3% for Qwen2-72B.
Side by side
| gpt-oss-120b | Qwen2-72B | |
|---|---|---|
| Provider | OpenAI | Alibaba (Qwen) |
| Noometry Index | 36.3 | 30.0 |
| Released | 2025-08-05 | 2024-06-07 |
| Weights | Open | Open |
| Context window | 131K | — |
| Max output | 41K | — |
| Input $ / M tokens | $0.037 | — |
| Output $ / M tokens | $0.17 | — |
| Results tracked | 48 | 26 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding gpt-oss-120b leads
gpt-oss-120b: 33.5 (#256), Qwen2-72B: 29.1 (#310)
| Benchmark | gpt-oss-120b | Qwen2-72B |
|---|---|---|
| WeirdML | 48.2% | 11.3% |
| LMArena Coding | 1380 | 1196 |
| SWE-bench Verified (bash only) | 26% | — |
| Aider Polyglot | 41.8% | — |
| SciCode | 36% | — |
| BigCodeBench Instruct | — | 38.5% |
| BigCodeBench Complete | — | 54% |
| ALE-Bench | 575.62 | — |
| AlgoTune | 1.41 | — |
Agentic & Tool Use Qwen2-72B leads
gpt-oss-120b: 12.2 (#153), Qwen2-72B: 17.0 (#146)
| Benchmark | gpt-oss-120b | Qwen2-72B |
|---|---|---|
| METR Time Horizons | 56.6% | 29.9% |
| Terminal-Bench | 18.7% | — |
| APEX-Agents | 4.4% | — |
| TheAgentCompany | — | 1.1% |
| Vending-Bench 2 | -21.53 | — |
Reasoning Qwen2-72B leads
gpt-oss-120b: 20.0 (#245), Qwen2-72B: 23.2 (#181)
| Benchmark | gpt-oss-120b | Qwen2-72B |
|---|---|---|
| LMArena Hard Prompts | 1364 | 1191 |
| Epoch Capabilities Index | 139.93 | 125.28 |
| SimpleBench | 22.1% | — |
| Kagi LLM Benchmark | 58.6% | — |
| CritPt | 1.1% | — |
| Chess Puzzles | 20% | — |
| Mystery Game Puzzles | 2% | — |
| DTBench | 76.3% | — |
| LMCA | 22.1% | — |
| Surface Evolver Bench | 25% | — |
Math gpt-oss-120b leads
gpt-oss-120b: 52.5 (#50), Qwen2-72B: 30.2 (#236)
| Benchmark | gpt-oss-120b | Qwen2-72B |
|---|---|---|
| LMArena Math | 1389 | 1235 |
| OTIS Mock AIME 2024-2025 | 88.9% | — |
| Omni-MATH | 68.8% | — |
| MATH Level 5 | — | 39.1% |
Knowledge gpt-oss-120b leads
gpt-oss-120b: 42.4 (#96), Qwen2-72B: 21.2 (#275)
| Benchmark | gpt-oss-120b | Qwen2-72B |
|---|---|---|
| GPQA Diamond | 75.8% | 40.8% |
| LMArena Expert | 1356 | 1171 |
| MMLU-Pro | 79.5% | — |
| Confabulations | 15.7% | — |
| Vectara Hallucination Rate | 14.2% | — |
| GPQA (HELM) | 68.4% | — |
| MMLU | — | 82.4% |
Multilingual gpt-oss-120b leads
gpt-oss-120b: 48.0 (#147), Qwen2-72B: 35.9 (#244)
| Benchmark | gpt-oss-120b | Qwen2-72B |
|---|---|---|
| LMArena Non-English | 1351 | 1176 |
| LMArena Chinese | 1385 | 1240 |
| LMArena French | 1369 | 1170 |
| LMArena German | 1353 | 1151 |
| LMArena Japanese | 1331 | 1111 |
| LMArena Korean | 1282 | 1083 |
| LMArena Russian | 1343 | 1169 |
| LMArena Spanish | 1389 | 1169 |
Instruction Following gpt-oss-120b leads
gpt-oss-120b: 69.3 (#173), Qwen2-72B: 61.7 (#241)
| Benchmark | gpt-oss-120b | Qwen2-72B |
|---|---|---|
| LMArena Instruction Following | 1318 | 1181 |
| IFEval | 83.6% | — |
Long Context Qwen2-72B leads
gpt-oss-120b: 31.4 (#278), Qwen2-72B: 36.1 (#235)
| Benchmark | gpt-oss-120b | Qwen2-72B |
|---|---|---|
| LMArena Longer Query | 1319 | 1192 |
| Fiction.LiveBench | 44.4% | — |
Writing & Preference gpt-oss-120b leads
gpt-oss-120b: 46.5 (#217), Qwen2-72B: 40.8 (#241)
| Benchmark | gpt-oss-120b | Qwen2-72B |
|---|---|---|
| LMArena Text | 1365 | 1203 |
| LMArena Creative Writing | 1275 | 1181 |
| LMArena Multi-Turn | 1340 | 1196 |
| Short-Story Creative Writing | 77.1% | — |
| EQ-Bench Creative Writing | 961 | — |
| WildBench | 84.5% | — |
Frequently asked questions
Is gpt-oss-120b better than Qwen2-72B?
gpt-oss-120b is the stronger model overall, scoring 36.3 to 30.0 on the Noometry Index.
Is gpt-oss-120b or Qwen2-72B better for coding?
gpt-oss-120b scores higher on coding benchmarks: 33.5 versus 29.1 in the Noometry coding category.
How many benchmarks do gpt-oss-120b and Qwen2-72B share?
21 benchmarks have published results for both models. gpt-oss-120b has 48 scored results on Noometry and Qwen2-72B has 26.