Bonus Inference,
Peak Rates &
Token Discounts
The current state of known inference bonuses or discount multipliers is shown in the table below. For periodic monitoring use our Prometheus-compatible /metrics endpoint.
| vendor | window | description | schedule | local now | active | changes in | event ends | docs |
|---|---|---|---|---|---|---|---|---|
| z.ai | peak | Time-variable rates apply during peak hours depending on plan and model. | weekly Mon,Tue,Wed,Thu,Fri 14:00–18:00 Asia/Shanghai | Sat 09:57 CST | 2.2d | docs ↗ | ||
| deepseek | peak | Peak pricing: off-peak rates are half of peak. New V4.1 Flash billing effective 2026-09-10 04:00 UTC. | weekly Mon,Tue,Wed,Thu,Fri 01:00–04:00 + 06:00–10:00 UTC | Sat 01:57 UTC | 47h 2m | docs ↗ | ||
| z.ai | flash | GLM-5.3-Flash bonus: unlimited Flash in ZCode, 2x quota in other supported agents. | 2026-09-03..2026-09-20 08:00–18:00 America/Los_Angeles | Fri 17:57 PST | over | 2026-09-21 01:00Z | docs ↗ |
z.aipeak
- description
- Time-variable rates apply during peak hours depending on plan and model.
- schedule
- weekly Mon,Tue,Wed,Thu,Fri 14:00–18:00 Asia/Shanghai
- local now
- Sat 09:57 CST
- changes in
- 2.2d
deepseekpeak
- description
- Peak pricing: off-peak rates are half of peak. New V4.1 Flash billing effective 2026-09-10 04:00 UTC.
- schedule
- weekly Mon,Tue,Wed,Thu,Fri 01:00–04:00 + 06:00–10:00 UTC
- local now
- Sat 01:57 UTC
- changes in
- 47h 2m
z.aiflash
- description
- GLM-5.3-Flash bonus: unlimited Flash in ZCode, 2x quota in other supported agents.
- schedule
- 2026-09-03..2026-09-20 08:00–18:00 America/Los_Angeles
- local now
- Fri 17:57 PST
- changes in
- over
- event ends
- 2026-09-21 01:00Z