Model profile
GLM-5.3
Z.aideveloper
2026-08-18release date
#26 / 317overall rank
3eval lineages
Evidence summary
GLM-5.3 has an estimated overall rank of #26; its 90% source-sensitivity interval is #10–#152. Its behavior-only rank is #11; company governance moves the combined estimate to #26. Published evidence spans 3 evals and 4 of 7 behavior components. Its strongest relative result is SM-Bench (anti_hallucination, #1 of 81); its weakest is TAC (base_welfare_rate, #28 of 80).
Compare this model
Only models sharing at least one published sub-eval are listed.
Official and reference links
- Artificial Analysis ↗glm-5-3
- OpenRouter ↗z-ai/glm-5.3
Published eval results
Rank is within that sub-eval. Black marks the observed result; the grey dot marks the value implied by the global rank. Values stay on each source’s native scale.
| Eval / sub-eval | Rank | Value | Distribution | Source |
|---|---|---|---|---|
| AA-Omnisciencehallucination_rate | #22 / 332 | ↓0.2955 | Source ↗official | |
| SM-Benchadversarial | #17 / 81 | ↑86.34 | Source ↗official | |
| SM-Benchambiguous_interpretation | #11 / 81 | ↑90.77 | Source ↗official | |
| SM-Benchanti_hallucination | #1 / 81 | ↑100 | Source ↗official | |
| SM-Bencheq_boundaries | #20 / 81 | ↑69.66 | Source ↗official | |
| SM-Benchoverfit | #8 / 81 | ↑92.35 | Source ↗official | |
| TACbase_welfare_rate | #28 / 80 | ↑29.49 | Source ↗self run |