Skip to content
NEWSee what changed →

zai-org · glm_moe_dsa

GLM-5.1

GLM-5.1 by zai-org: a 753.9B-parameter mixture-of-experts open-weight model under the mit license.

mitMixture of experts203K contextAuto-imported · not yet reviewed

742,272 downloads · 1,851 likes · Model card · synced 2026-11-19

Specifications

Parameters
753.9B (23.6B active)
Architecture
moe · glm_moe_dsa
Layers
78
Hidden size
6,144
KV heads · head dim
64 · 64
Vocabulary
154,880
Native dtype
—
Context window
202,752 tokens
Released
2026-04-03
Last modified on Hub
2026-05-13

Features & licensing

License
mit
Commercial use
Yes
Modalities
text
Tool / function calling
No
Reasoning mode
No
Languages
EN, ZH
Quantised variants on Hub
AWQ, FP8, GGUF, bitsandbytes
Pipeline
text-generation
transformerssafetensorsglm_moe_dsatext-generationconversationalenzheval-resultsendpoints_compatible

Task fit

Editorial scores (0–100) used by the recommendation engine.

  • Chat70
  • RAG / Q&A70
  • Code70
  • Summarisation70
  • Extraction70
  • Agents70

Hardware to run GLM-5.1

Weights need about 1659 GB at FP16, 871 GB at INT8 and 498 GB at INT4. The KV cache adds roughly 1278 MB per 1,000 tokens per request. For a reference workload of 5 requests per second with 1,500 input and 300 output tokens, total GPU memory is around 1116 GB.

GPU configurations at the reference workload
ConfigurationVRAMUtilisationEst. first tokenCloud / month

Estimates only. Run an assessment for your own traffic.

Auto-imported from Hugging Face. Architecture, license and feature data are synced from the model repository; the task-fit scores below are provisional defaults until a curator reviews them.