Skip to content

Relace models

Get API keyRead the docs
Relace is a hosted inference provider for open-weight AI models, delivering cost-efficient access optimized for coding, reasoning, and agentic workloads.
Model
Context
Input
Output
Latency
Throughput
Cache
Web Search
Capabilities
ZDR
No Training
Regional Inference
Free Tier
Released
zai/glm-5.3-flash
1M
$0.07/M
$0.24/M
——
Read$0.01/M
—
+1
08/26/2026
deepseek/deepseek-v4-flash-0731
1M
$0.05/M
$0.10/M
1.0 s116 tps
Read$0.01/M
—
07/31/2026