Across 16 shared benchmarks, Kimi K2.6 scores higher on 11 and muse-glimmer-30b on 5. The widest gap is HLE, where Kimi K2.6 scores 37.5 against 22. muse-glimmer-30b is the cheaper of the two on tracked API pricing ($0.30 against $0.91 per million input tokens).
| Benchmark | Kimi K2.6 | muse-glimmer-30b |
|---|---|---|
| AA-LCR | 76.7 | 80 |
| aime_2026 | 96.4 | 94.7 |
| baby_vision | 68.5 | 70.4 |
| charxiv_rq | 86.7 | 85.9 |
| deepsearchqa_f1 | 92.5 | 74.6 |
| GPQA Diamond | 91.1 | 83.5 |
| HLE | 37.5 | 22 |
| IFBench | 76 | 77 |
| MCP Atlas | 68.1 | 75.5 |
| MMMU-Pro | 80.1 | 74 |
| OSWorld-Verified | 73.1 | 65.9 |
| scicode | 53.5 | 43.6 |
| SWE-bench Pro | 58.6 | 51.2 |
| SWE-bench Verified | 80.2 | 76 |
| Tau 3 Banking | 20.6 | 23.5 |
| Terminal-Bench 2.1 | 65.9 | 51.7 |
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 (moonshot-official, deepinfra), otherwise the lowest tracked offer.