DeepSeek V4 Flash 0731 repriced across 3 hosts: NextBit output down 14%, Relace output up 100%
Published 21 September 2026
Price move, per 1M tokens
- Relace · inputunchanged$0.040 → $0.040
- Relace · output↑ 100%$0.16 → $0.32
- Relace · cache readunchanged$0.016 → $0.016
- OpenInference · inputunchanged$0.040 → $0.040
- OpenInference · output↑ 18%$0.17 → $0.20
- OpenInference · cache readunchanged$0.010 → $0.010
- NextBit · inputunchanged$0.15 → $0.15
- NextBit · output↓ 14%$0.35 → $0.30
- NextBit · cache readunchanged$0.035 → $0.035
| host | rate | was | now | change |
|---|---|---|---|---|
| Relace | input | $0.040 | $0.040 | — |
| Relace | output | $0.16 | $0.32 | ↑ 100% |
| Relace | cache read | $0.016 | $0.016 | — |
| OpenInference | input | $0.040 | $0.040 | — |
| OpenInference | output | $0.17 | $0.20 | ↑ 18% |
| OpenInference | cache read | $0.010 | $0.010 | — |
| NextBit | input | $0.15 | $0.15 | — |
| NextBit | output | $0.35 | $0.30 | ↓ 14% |
| NextBit | cache read | $0.035 | $0.035 | — |
Relace output: ↑ 100%OpenInference output: ↑ 18%NextBit output: ↓ 14%
The model
DeepSeek V4 Flashopen the model →
Open weights290.9B parameters1049K context33 hosts
intelligence65th of 168 writing57th of 168 coding64th of 168 agents20th of 55
DeepSeek V4 Flash is a downloadable text model with a permissive licence, built for long-document work where the bill matters more than peak quality. Its measured quality is weak on the LiveBench boards and mid-pack on the Arena boards, so treat it as a cheap workhorse rather than a quality leader.
Source
Read on openrouter.ai · we published this on 21 September 2026.
read the original at openrouter.ai ↗