Back to models
DeepSeek · DeepSeek

DeepSeek V4 Pro

deepseek-v4-pro

DeepSeek's open-weight V4 flagship for deep reasoning, coding and long-horizon agent tasks.

Compare models
Input and output modalities
Input/output price
$0.693 / $2.077 / 1M token
Context window
1M
Providers
5
Release date
Aug 13, 2026

Providers

Effective prices, cache hit rate, and live availability for every provider.

Alibaba Cloud
$0.693 / $2.077
-
-
Baidu Cloud
$0.693 / $2.077
-
-
DeepSeek
$0.693 / $2.077
-
-
Tencent Cloud
$0.693 / $2.077
-
-
Volcengine
$1.385 / $4.154
-
-

Effective prices include each mapping's discount. Billing uses the provider mapping actually selected for the request.

Performance

DDSketch latency percentiles by minute or hour, defaulting to P95.

Latency

Time to first token

Availability

Cache hit rate

Applications

Client-reported applications ranked by token usage over the last 24 hours.

Last 24 hours
No application source data