zai
GLM 5.3 Fast
Speed-optimized version of Z.AI’s GLM-5.3 agentic coding model, built for responsive, real-time workloads. It supports advanced coding, long-horizon task execution, testing, vulnerability discovery, and cybersecurity analysis, making it ideal for developer tools, coding agents, and complex multi-step engineering workflows.
reasoningtool useimplicit cachingstructured outputtext
Overview
- Type
- language
- Context
- 1M
- Max output
- 262K
- Input / output
- text / text
API
Model identifier
zai/glm-5.3-fastProviders and live performance
Rolling Gateway metrics. Latency is time to first token; throughput is output tokens per second.
| Provider | Status | Latency | Throughput | Uptime 15m / 1h / 1d | Input | Output | Cache read / write | Discount | Regions | ZDR | No training |
|---|---|---|---|---|---|---|---|---|---|---|---|
| alibaba | Operational | 1.65s / 3.15s | 82.5 / 139.0 tok/s | 100.00% / 100.00% / 100.00% | $2.80/M | $8.80/M | $0.56/M / — | — | — | Yes | Yes |
| baseten | Operational | 0.70s / 2.85s | 345.0 / 446.1 tok/s | 100.00% / 100.00% / 99.98% | $2.10/M | $6.60/M | $0.21/M / — | — | us | Yes | Yes |
Pricing
input
0.0000021
output
0.0000066
input cache read
0.00000021
varies by provider
Yes
regional
{
"us": {
"input": "0.0000021",
"output": "0.0000066",
"input_cache_read": "0.00000021"
}
}
Capabilities
reasoningtool useimplicit cachingstructured outputtext
Supported parameters
max tokens
Yes
temperature
Yes
stop
Yes
tools
Yes
tool choice
Yes
reasoning
Yes
include reasoning
Yes
response format
Yes
structured outputs
Yes
Reasoning options
options
{
"type": "toggle"
}, {
"type": "effort",
"values": [
"none",
"low",
"medium",
"high"
]
}
Metadata
type
language
context window
1M
max output
262K
released
Sep 2, 2026
temperature control
Yes
zero data retention
all
no training
all
regions
us
api specifications
v2, v3, v4