Rate Limits & Quotas
Prompt limits per plan and the 5-hour window. All model variants are FP8-quantised.
Drise uses prompt-based quotas per 5-hour rolling window. Limits reset every 5 hours. All five GLM 5.2 model variants (see Models) share the same per-plan quota.
Per-plan limits
| Plan | Price | Prompts per 5 hours |
|---|---|---|
| Free | $0 | 30 |
| Pro | $15/mo | 1,500 |
| Max | $25/mo | 3,000 |
| Turbo | $50/mo | 5,000 |
Model-specific notes
drise-glm-5.2anddrise-glm-5.2-fastexpose a 1,000,000-token context window.drise-glm-5.2-shortanddrise-glm-5.2-short-fastuse a 200,000-token context window for focused tasks.drise-visionaccepts image and text inputs.- All variants are FP8-quantised for cheaper, faster inference.
Notes
- Higher tiers get higher routing priority. See Routing & Failover.
- No per-token math. No surprise overages.
- Upgrade or downgrade anytime, no lock-in.