DeepSeek V4 Flash
DeepSeek
🧠 Language Model
DeepSeek V4 Flash, fast, cost-effective daily driver. 1M context, thinking + non-thinking modes.
Specifications
Input modalitiestext
Output modalitiestext
Capabilitieschat, reasoning, tools
Relative speed●●●●● Blazing
Context window1,000,000 tokens
Max output65,536 tokens
Tool callingYes
Input price$0.385 / 1M
Output price$1.16 / 1M
Pricing
$0.385 / 1M input tokens · $1.16 / 1M output tokens
Platform credit pricing — live rates on the Models page.