GLM 5.3 repriced across 2 hosts: Inceptron input down 22%, Wafer cache read up 11%
Published 23 September 2026
Price move, per 1M tokens
- Inceptron · input↓ 22%$1.01 → $0.79
- Inceptron · output↓ 19%$3.29 → $2.65
- Inceptron · cache read↓ 20%$0.255 → $0.203
- Wafer · input↓ 17%$1.19 → $0.99
- Wafer · outputunchanged$4.40 → $4.40
- Wafer · cache read↑ 11%$0.36 → $0.40
| host | rate | was | now | change |
|---|---|---|---|---|
| Inceptron | input | $1.01 | $0.79 | ↓ 22% |
| Inceptron | output | $3.29 | $2.65 | ↓ 19% |
| Inceptron | cache read | $0.255 | $0.203 | ↓ 20% |
| Wafer | input | $1.19 | $0.99 | ↓ 17% |
| Wafer | output | $4.40 | $4.40 | — |
| Wafer | cache read | $0.36 | $0.40 | ↑ 11% |
Inceptron input: ↓ 22%Wafer input: ↓ 17%Wafer cache read: ↑ 11%
If you buy
Inceptron now suits steady, input-heavy work better than before, so re-check your cached-token share and whether your usage still fits this host's mix.
The model
GLM 5.3open the model →
Proprietary1049K context38 hosts
intelligence16th of 168 writing19th of 168 coding23rd of 168 agents17th of 55
GLM 5.3 is a hosted-only text model: we list no download for it, so using it means choosing a host. It is strongest where people rate the answers themselves, and mid-field on the harder task sets.
Source
Read on openrouter.ai · we published this on 23 September 2026.
read the original at openrouter.ai ↗