GLM 5.3 Prime
Z AI · 1 configuration
Intelligence—
Input / 1M$2.80
Output / 1M$8.80
Context1M
Tariff: OpenRouter · Source: openrouter · USD per million tokens
Thinking configuration
What does more thinking buy you?
Compare this model’s measured thinking configurations.
◌
No pair of measurements available
Change the filters or pick another benchmark. Missing values are not estimated.
Selected: thinking default · Click a point to switch configuration
GLM-5.3-Prime is the high-speed variant of Z.ai's GLM-5.3, inheriting its full capabilities while delivering 1.5–2× the output throughput through inference acceleration. It supports text input and output with a 1M-token...
This model is in the catalog but has no imported benchmarks. Capabilities are not estimated.
Specifications from the catalog
Inputtext
Outputtext
Tool callingYes
Structured outputNo
SpeedNot available
First tokenNot available
Maximum output131.1K
Catalog listing date2026-09-23