Z.ai: GLM 5.3 Prime

Name: Z.ai: GLM 5.3 Prime
Author: OpenRouter

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.

Context

Max output

131K

Input price

$2.8/1M tokens

Output price

$8.8/1M tokens

Capabilities

Show unsupported (4)

Reasoning & Interaction

Reasoning

Tool Calling

Output Control

Structured Output

JSON Mode

Details

Model IDz-ai/glm-5.3-prime

ProviderOpenRouter

Creatorz-ai

Familyglm-5.3

License—

Status—

Typechat

Input modalitiestext

Output modalitiestext

Open weight—

Parameters—

Architecture—

Context window1M tokens

Max input—

Max output131K tokens

Reasoning tokens—

Knowledge cutoff—

Release date2026-09-23

Sourceofficial

Tools

Function Callingfunction_calling

Call external functions and APIs

Pricing

Input

$2.8

Output

$8.8

Cache write

—

Cache read

$0.56

Batch in

—

Batch out

—

Family Comparison: glm-5.3

Model	Caps	Perf	Speed	Context	Max out	Pricing
Z.ai: GLM 5.3 Flash	RTJJM	—	—	1.0M	131K	$0.05$0.07
Z.ai: GLM 5.3 Flash (batch)	RTJJM	—	—	1.0M	131K	$0.05$0.18
Z.ai: GLM 5.3	RTJJM	—	—	1.0M	944K	$0.08$3.6
Z.ai: GLM 5.3 FlashX	RTJJM	—	—	1.0M	131K	$0.37$1.25
Z.ai: GLM 5.3 (batch)	RTJJM	—	—	1.0M	131K	$0.45$2
Z.ai: GLM 5.3 Prime	RTJJM	—	—	1M	131K	$2.8$8.8

API

GET/v1/models/openrouter/z-ai/glm-5.3-prime