GLM-5.3

Zhipu AI🇨🇳 China
active
Context window1049K tokens
Input / 1M tokens$1.4
Output / 1M tokens$4.4

Version History

5.3minor

GLM-5.3 uses the same base model as GLM-5.2 with extended post-training, adding cybersecurity-focused training data and improving agentic coding task performance. Weights are set to open source in about two weeks pending security review.

Benchmark Scores

Full leaderboard →
91.7%
GPQA
84.0 tokens_per_sec
Speed (tok/s)

Coverage