Zhipu AI·released 2025-09-30
Scored on 4 benchmarks. 50% of its scores are independently reproduced — the rest are self-reported or unverified.
Every score, ranked, colored by evidence status.
| Benchmark | Score | Evidence status | Measured |
|---|---|---|---|
| Terminal Bench | 24.5accuracy (%) | unverified· optimized | 2025-09-30 |
| FrontierMath-2025-02-28-Private | 6.7accuracy (%) | reproduced· optimized | 2025-09-30 |
| FrontierMath-Tier-4-2025-07-01-Private | 3.55accuracy (%) | reproduced | 2025-09-30 |
| APEX-Agents | 3accuracy (%) | unverified | 2025-09-30 |