GLM-4.6

Z.ai llm

Z.ai's successor to GLM-4.5, expanding context to 200K tokens and improving coding and agentic performance. Z.ai publishes its benchmark results only as chart images, so no sourced scores are recorded here yet.

overall ⌄

34.1

108 of 125 ranked

price · $/M tokens

0.6 / 2.2

in / out, as of 2026-07-25

to run it yourself ⌄

214 GB

minimum · 857 GB recommended

Benchmarks

What its maker published, and what anyone else measured. Every figure links the document it came from.

benchmarkpublishedmeasured
SWE-bench68 55.4
HLE17.2 —
BrowseComp45.1 —

Cheaper, and at least as good

On a 1,000-in, 1,000-out call these cost less and rank no lower — a hard comparison to argue with, and a narrow one. Value covers what it misses.

Devstral Small 2 Mistral AI$0.00040overall 44.3
Ministral 3 14B Reasoning Mistral AI$0.00040overall 47.5
DeepSeek V4-Flash DeepSeek$0.00042overall 57.9
Claude Haiku 5.5 Anthropic$0.00060overall 65.3
Qwen3.8-Flash-Next Alibaba$0.00062overall 61.2

Lineage

↑ finetune GLM-4.5
view in full graph →

Metadata

orgZ.ai
released2025-09-30
params357B-A32B
licensemit
hf idzai-org/GLM-4.6