Head to head
GLM 5.3 leads on 7 of 11 shared benchmarks.
GLM 5.3 leads on 7 of 11 shared benchmarks.
| Dimension | GLM 5.3 | Kimi K3 |
|---|---|---|
| GPQA Diamond | 88.1% | 92.9% |
| HLE | 42% | 43.5% |
| MathArena Apex | — | 65.6% |
| Terminal-Bench 2.1 | 88.2% | 88.3% |
| Terminal-Bench 3.0 | 28.3% | 17.7% |
| Terminal-Bench 4.0 | 37.9% | 12.6% |
| DeepSWE v1.1 | 66.9% | 67.5% |
| ProgramBench | 19% | 17.5% |
| NL2Repo-Bench | 58% | 58% |
| CyberGym | 84.5% | 80% |
| ExploitGym | 15% | — |
| HLE (w/tools) | 62.5% | 59.8% |
| Automation-Bench | 48.8% | 46.7% |
| Agents' Last Exam | 28.5% | 27.6% |
| Chartography (w/tools) | — | 68.1% |
| BabyVision (w/tools) | — | 85.7% |
| ZeroBench-main (w/tools) | — | 41% |
| Blended price | — | — |
| Full priceInput · Output · Cached input | — | — |
| Context window | — | — |
Adjust the monthly volume — every model recomputes instantly.
Estimates use list prices and ignore cached input, batch discounts, and off-peak rates.