GLM-5

Open weights Active Text LLM MoE

Open 744B-parameter MoE model with DeepSeek Sparse Attention, aimed at agentic engineering.

Released
Feb 12, 2026
Context window
200K tokens
Max output
131K tokens
Parameters
744B 40B active
Input price
$1.00 per 1M
Output price
$3.20 per 1M

Prices are first-party list prices in USD at the standard tier. Last verified Oct 30, 2026.

Pricing

Launch priceFeb 12, 2026
$1.00 in / $3.20 out
Current price since Feb 12, 2026
$1.00 in / $3.20 out
Cached input $0.20 per 1M

Specifications

Developer
Z.ai (Zhipu AI)
Model type
Text LLM
Architecture
MoE
Parameters
744B (40B active)
Input
Text
Output
Text
Knowledge cutoff
Not disclosed
License
MIT License
API access
First-party API
Capabilities
Agentic Coding Flagship Reasoning