DeepSeek V4 Pro 0813 repriced across 3 hosts: Sail Research (US region) input down 43%, output up 45%

Published 22 September 2026

Price move, per 1M tokens

Ionstream · inputunchanged
$0.62 → $0.62
Ionstream · outputunchanged
$2.88 → $2.88
Ionstream · cache read↓ 50%
$0.088 → $0.044
Sail Research (US region) · input↓ 43%
$0.70 → $0.40
Sail Research (US region) · output↑ 45%
$2.96 → $4.30
Sail Research (US region) · cache readunchanged
$0.033 → $0.033
Sail Research · input↓ 43%
$0.70 → $0.40
Sail Research · output↑ 45%
$2.96 → $4.30
Sail Research · cache readunchanged
$0.033 → $0.033
Wafer · input↓ 25%
$0.67 → $0.50
Wafer · outputunchanged
$2.90 → $2.90
Wafer · cache readunchanged
$0.16 → $0.16

host	rate	was	now	change
Ionstream	input	$0.62	$0.62	—
Ionstream	output	$2.88	$2.88	—
Ionstream	cache read	$0.088	$0.044	↓ 50%
Sail Research (US region)	input	$0.70	$0.40	↓ 43%
Sail Research (US region)	output	$2.96	$4.30	↑ 45%
Sail Research (US region)	cache read	$0.033	$0.033	—
Sail Research	input	$0.70	$0.40	↓ 43%
Sail Research	output	$2.96	$4.30	↑ 45%
Sail Research	cache read	$0.033	$0.033	—
Wafer	input	$0.67	$0.50	↓ 25%
Wafer	output	$2.90	$2.90	—
Wafer	cache read	$0.16	$0.16	—

Ionstream cache read: ↓ 50%Sail Research (US region) input: ↓ 43%Sail Research (US region) output: ↑ 45%Sail Research input: ↓ 43%Sail Research output: ↑ 45%Wafer input: ↓ 25%

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 22 September 2026.

← all news