DeepSeek V4 Pro repriced across 4 hosts: Io Net input down 11%, Relace input up 72%
Published 28 September 2026
Price move, per 1M tokens
- Relace · input↑ 72%$0.204 → $0.350
- Relace · output↓ 7%$3.78 → $3.50
- Relace · cache read↓ 51%$0.203 → $0.100
- Wafer · input↑ 35%$0.293 → $0.397
- Wafer · output↑ 20%$3.50 → $4.20
- Wafer · cache read↑ 85%$0.172 → $0.318
- Io Net · input↓ 11%$1.24 → $1.10
- Io Net · outputunchanged$3.50 → $3.50
- Io Net · cache read↓ 14%$0.14 → $0.12
- Ionstream · input↓ 11%$0.253 → $0.226
- Ionstream · outputunchanged$1.96 → $1.96
- Ionstream · cache readunchanged$0.088 → $0.088
| host | rate | was | now | change |
|---|---|---|---|---|
| Relace | input | $0.204 | $0.350 | ↑ 72% |
| Relace | output | $3.78 | $3.50 | ↓ 7% |
| Relace | cache read | $0.203 | $0.100 | ↓ 51% |
| Wafer | input | $0.293 | $0.397 | ↑ 35% |
| Wafer | output | $3.50 | $4.20 | ↑ 20% |
| Wafer | cache read | $0.172 | $0.318 | ↑ 85% |
| Io Net | input | $1.24 | $1.10 | ↓ 11% |
| Io Net | output | $3.50 | $3.50 | — |
| Io Net | cache read | $0.14 | $0.12 | ↓ 14% |
| Ionstream | input | $0.253 | $0.226 | ↓ 11% |
| Ionstream | output | $1.96 | $1.96 | — |
| Ionstream | cache read | $0.088 | $0.088 | — |
Relace output: ↓ 7%Relace input: ↑ 72%Wafer input: ↑ 35%Io Net input: ↓ 11%Ionstream input: ↓ 11%
If you buy
Relace now charges more for fresh input but less for output and cached reads, so prompt-reuse and long-generation workloads gain while input-heavy jobs should re-check the bill.
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 on openrouter.ai · we published this on 28 September 2026.
read the original at openrouter.ai ↗