models · zhipu-ai
Zhipu AI GLM-5. 744B MoE, 200K context, 351 tok/s. Arena score 1593. Strong multilingual reasoning and instruction following.
| Weight file SHA-256 | 4443c1888927bd19c3728a10043a1cd9a2369430efef192ffa70fc22f7be12c2 |
|---|---|
| Method | hf-published-sha256 |
| Publisher | zhipu-ai |
|---|---|
| Origin | Recorded as a covered nation under 10 U.S.C. § 4872(f) — the PRC, Russia, Iran or North Korea. The registry did not record which. |
| Licence | MIT |
| Parameters | 753,864,139,008 (753.9B) |
| Context window | 200K |
| Modality | text->text |
| Architecture | glm_moe_dsa |
| Layers | 78 |
| Hidden size | 6144 |
| Attention | MLA (64 heads, 64 KV) |
| Native precision | BF16 |
| Pulls recorded here | 480,016 |
| Downloads reported upstream | 56,846 |
| GPQA Diamond | 82 |
|---|---|
| IFBench | 72.3 |
| Humanity's Last Exam | 29.3 |
Reported by a third party and recorded here, not re-run by us. Source: artificialanalysis.ai/api/v2.
From local-ai-registry (MIT), commit 124959d: 18 runs across 2 machines, 0 validated — model revision and runtime pinned, launch accepted — and 18 candidate, their label for useful evidence without a reproducible-launch promise.
Registry record as of 2026-09-26. Identity, licence and parameter count are properties of a release and do not move; tier can, and the verify link above re-checks it against the live registry rather than this page.