Across 20 shared benchmarks, GLM-5.2 Full Open Source scores higher on 18 and Qwen3.5 122B A10B on 2. The widest gap is OmniScience Non-Hallucination, where GLM-5.2 Full Open Source scores 73.7 against 12.9. Qwen3.5 122B A10B is the cheaper of the two on tracked API pricing ($0.40 against $1.40 per million input tokens).
| Benchmark | GLM-5.2 Full Open Source | Qwen3.5 122B A10B |
|---|---|---|
| AA Agentic Index | 45.7 | 21.3 |
| AA Intelligence | 52.6 | 32.8 |
| AA-LCR | 76.7 | 70.3 |
| AA-Omniscience | 4.4 | -41.5 |
| Artificial Analysis Coding Index | 68.8 | 45.7 |
| coding_arena_elo | 1582 | 1358 |
| critpt | 20.9 | 0.9 |
| gdpval | 50.3 | 24.3 |
| GPQA Diamond | 91.2 | 86.6 |
| HLE | 54.7 | 47.5 |
| HMMT 2025 | 94.4 | 90.3 |
| IFBench | 73.3 | 76.1 |
| OmniScience Accuracy | 24.3 | 24.4 |
| OmniScience Non-Hallucination | 73.7 | 12.9 |
| scicode | 50.5 | 42 |
| TauBench V3 - Banking | 34.6 | 15.3 |
| Terminal-Bench 2.1 | 82.7 | 47.6 |
| Terminal-Bench Hard | 50.8 | 31.1 |
| τ²-Bench Telecom (AA run) | 99.1 | 93.6 |
| τ³-Bench | 26.8 | 13.6 |
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 (zai-official, alibaba-official), otherwise the lowest tracked offer.