DeepSeek V4 Pro repriced across 2 hosts: Wafer input down 36%, output up 21%
Published 25 September 2026
Price move, per 1M tokens
- Wafer · input↓ 36%$0.39 → $0.25
- Wafer · output↑ 21%$2.90 → $3.50
- Wafer · cache read↓ 0.4%$0.250 → $0.249
- Ionstream · input↓ 35%$0.390 → $0.253
- Ionstream · output↓ 32%$2.88 → $1.96
- Ionstream · cache readunchanged$0.088 → $0.088
| host | rate | was | now | change |
|---|---|---|---|---|
| Wafer | input | $0.39 | $0.25 | ↓ 36% |
| Wafer | output | $2.90 | $3.50 | ↑ 21% |
| Wafer | cache read | $0.250 | $0.249 | ↓ 0.4% |
| Ionstream | input | $0.390 | $0.253 | ↓ 35% |
| Ionstream | output | $2.88 | $1.96 | ↓ 32% |
| Ionstream | cache read | $0.088 | $0.088 | — |
Wafer input: ↓ 36%Wafer output: ↑ 21%Ionstream input: ↓ 35%
If you buy
On Ionstream the balance shifts toward prompt-heavy work, so re-check your output share before your next top-up.
The model
DeepSeek V4 Proopen the model →
Open weights1598.8B parameters1049K context28 hosts
intelligence36th of 168 writing32nd of 168 coding43rd of 168 agents31st of 55
DeepSeek V4 Pro is a large downloadable text model with a one-million-token request limit and strong mathematics scores. Its mixture-of-experts design keeps only 37 billion parameters active per token, making it more efficient to run than its total size suggests.
Source
Read from the hosts' own published rates · we published this on 25 September 2026.