DeepSeek V4 Pro
DeepSeekdeepseek-v4-pro
Released Apr 24, 2026
Providers
5
Context
1.0M
Input
$0.50/M
Output
$2.00/M
Aliases:deepseek-prods-v4-pro
DeepSeek's flagship reasoning model with 1M context window. Supports tool calling, extended thinking, and high-quality generation.
relace/deepseek-v4-proQuantfp4
- Input
- $0.50/M
- Output
- $2.00/M
- Context
- 1.0M
- Max out
- 393K
- Cache read
- $0.05/M
- Cache write
- $0.50/M
- Thinking out
- $2.00/M
FlexPriority / Fast
alibaba/deepseek-v4-proQuantUnspecified
- Input
- $1.65/M
- Output
- $3.30/M
- Context
- 1.0M
- Max out
- 393K
- Cache read
- $0.14/M
- Cache write
- $1.65/M
- Thinking out
- $3.30/M
FlexPriority / Fast
deepinfra/deepseek-v4-proQuantfp8
- Input
- $1.30/M
- Output
- $2.60/M
- Context
- 1.0M
- Max out
- 33K
- Cache read
- $0.10/M
- Cache write
- $1.30/M
- Thinking out
- $2.60/M
nebius/deepseek-v4-proQuantfp8
- Input
- $1.75/M
- Output
- $3.50/M
- Context
- 1.0M
- Max out
- 393K
- Caching
- Not offered
- Thinking out
- $3.50/M
FlexPriority / Fast
novita/deepseek-v4-proQuantfp8
- Input
- $1.60/M
- Output
- $3.20/M
- Context
- 1.0M
- Max out
- 393K
- Cache read
- $0.14/M
- Cache write
- $1.60/M
- Thinking out
- $3.20/M
FlexPriority / Fast