Across 19 shared benchmarks, Gemini 3.6 Flash scores higher on 18 and Qwen3.5 122B A10B on 1. The widest gap is OmniScience Non-Hallucination, where Gemini 3.6 Flash scores 44.4 against 12.9. Qwen3.5 122B A10B is the cheaper of the two on tracked API pricing ($0.40 against $0.75 per million input tokens).
| Benchmark | Gemini 3.6 Flash | Qwen3.5 122B A10B |
|---|---|---|
| AA Agentic Index | 40.5 | 21.3 |
| AA Intelligence | 52 | 32.8 |
| AA-LCR | 79 | 70.3 |
| AA-Omniscience | 22.1 | -41.5 |
| arena_vision | 1302 | 1246 |
| Artificial Analysis Coding Index | 69.2 | 45.7 |
| coding_arena_elo | 1539 | 1358 |
| critpt | 10.6 | 0.9 |
| gdpval | 46 | 24.3 |
| GPQA Diamond | 92.8 | 86.6 |
| HLE | 40.8 | 47.5 |
| LVBench | 84.2 | 74.4 |
| MMMU-Pro | 83.2 | 76.9 |
| OmniScience Accuracy | 50 | 24.4 |
| OmniScience Non-Hallucination | 44.4 | 12.9 |
| OSWorld-Verified | 83 | 58 |
| scicode | 52.7 | 42 |
| TauBench V3 - Banking | 29.9 | 15.3 |
| Terminal-Bench 2.1 | 78 | 47.6 |
Best tracked score per model per benchmark (default configuration; source-attributed). ↓ marks lower-is-better metrics. Open either model for its full surface, provenance and pricing. Quoted rates are the price-setter row we currently track for each model — its direct or vendor-official listing where one exists (direct, direct), otherwise the lowest tracked offer.