Skip to content

GLM-4.6

Released 2025-09 · hybrid · 200K tokens · 4 benchmarks · Open weight

Editorial notes

MoE 357B / 32B activos. Cerca de paridad con Claude Sonnet 4 (48.6% win-rate CC-Bench).

Spec sheet

Company
Zhipu AI
Country
CN
Type
hybrid
Release
2025-09
Context
200K tokens
Params total
357B
Params active (MoE)
32B
License
mit
Quants
BF16, Q8_0, Q5_K_M, Q4_K_M, Q3_K_M
Slug
glm-4-6

Benchmarks (4)

Cite this model
BibTeX · APA

BibTeX

@misc{frontier-glm-4-6,
  title  = {GLM-4.6},
  author = {{Zhipu AI}},
  year   = {2025},
  note   = {Frontier Benchmarks AI atlas. Accessed 2026-05-08},
  url    = {https://frontierbenchmarks.com/models/glm-4-6}
}

APA

Zhipu AI (2025). GLM-4.6 [Large language model]. Frontier Benchmarks AI. Retrieved 2026-05-08, from https://frontierbenchmarks.com/models/glm-4-6

Citation reflects the atlas page, not the original model paper. For the paper, see the "Resources" section above.