DeepSeek V4.1 Flash cut across 2 hosts, by up to 47% at Sail Research (output)
Published 26 September 2026
Price move, per 1M tokens
- Sail Research · input↓ 38%$0.130 → $0.080
- Sail Research · output↓ 47%$0.75 → $0.40
- Sail Research · cache readunchanged$0.010 → $0.010
- Relace · input↓ 44%$0.090 → $0.050
- Relace · output↓ 11%$0.45 → $0.40
- Relace · cache read↓ 44%$0.009 → $0.005
| host | rate | was | now | change |
|---|---|---|---|---|
| Sail Research | input | $0.130 | $0.080 | ↓ 38% |
| Sail Research | output | $0.75 | $0.40 | ↓ 47% |
| Sail Research | cache read | $0.010 | $0.010 | — |
| Relace | input | $0.090 | $0.050 | ↓ 44% |
| Relace | output | $0.45 | $0.40 | ↓ 11% |
| Relace | cache read | $0.009 | $0.005 | ↓ 44% |
Sail Research output: ↓ 47%Relace input and cache read: ↓ 44%
If you buy
Output-heavy work at Sail Research now costs less to run, so re-check your usage split and whether your current plan still matches how you actually call this model.
The model
DeepSeek V4.1 Flashopen the model →
Open weights763.2B parameters1049K context35 hosts
intelligence21st of 168 writing37th of 168 coding11th of 168 agents13th of 55
DeepSeek V4.1 Flash is a downloadable model with a permissive licence, and it places 11th of 168 on Arena Coding via Max as of 25 Sep 2026. At 763.2 billion parameters it is a data-centre job to run yourself, so hosted use is the practical route for most readers.
Source
Read from the hosts' own published rates · we published this on 26 September 2026.