benchrInput / 1M$1.40hosted API
Output / 1M$4.40hosted API
Cached / 1M$0.26storage temporarily free
Context1M128K max output
Official price table
| Tier | Input / 1M | Output / 1M | Cached input / 1M |
|---|---|---|---|
| Hosted API | $1.40 | $4.40 | $0.26 |
Hosted price is not a self-hosting quote
Z.AI released the hosted endpoint and announced that weights would follow after safety hardening. Until the official artifact appears, do not model a self-hosted cluster as an available alternative. Reasoning is always enabled and defaults to max, so measure completion volume under the intended effort.
Frequently asked questions
How much does GLM-5.3 cost?
$1.40 input, $4.40 output, and $0.26 cached input per million tokens.
Is cache storage free?
The pricing page labels cache storage limited-time free; that can change independently of cache-hit tokens.
Are the weights available?
The API is live. Z.AI announced a later weight release after a safety-hardening period; verify the official artifact before self-hosting.