GLM-4.7 glm-4.7

zai-org · glm hybrid-reasoning flagship agentic

unranked — no quality signals yet

params
358B
arch
moe
context
198k
license
mit open weights
released
Dec 2025
reasoning
yes
train compute
4.4e+24 FLOP
downloads/30d
51.1k

# trend

collecting history — trends appear after a few daily runs.

# benchmarks

benchmarkscoresourcedate
AIME 2025 1.0 / 1 LLM Stats

# where to run

providerquantctx$/M in$/M out$/M blendtpsuptime
DeepInfrafp4 198k $0.40$1.75$0.74 99.9%
StreamLakefp8 200k $0.48$1.76$0.80 96.2%
AtlasCloudfp8 198k $0.52$1.85$0.85 98.6%
Novitafp8 200k $0.54$1.98$0.90 97.1%
Googleunknown 200k $0.60$2.20$1.00 100.0%
Z.AIfp4 198k $0.60$2.20$1.00 94.3%
Venicefp4 198k $0.55$2.65$1.07 100.0%
Phalaunknown 128k $0.85$3.30$1.46 99.5%
Cerebrasfp16 128k $2.25$2.75$2.38 100.0%

blended = (3·input + output) / 4 per 1M tokens · cheapest first

source aliases
aa
glm-4.7
epoch
GLM-4.7
openrouter
z-ai/glm-4.7