DeepSeek V4 Flash 0731

DeepSeek V4 Flash 0731 is a sparse mixture-of-experts model from DeepSeek, with 13B active parameters out of 284B total. This re-post-trained revision is suited for coding, reasoning, and agent workflows....

Released

Context window

1M tokens

Max output

944K tokens

Input

text

Output

text

Reasoning

Yes

Price history

• input $0.44 → $0.119 -73% · output $1.32 → $0.238 -82%
• input $0.044 → $0.44 +891% · output $0.133 → $1.32 +890%
• input $0.047 → $0.044 -6% · output $0.141 → $0.133 -6%
• input $0.44 → $0.047 -89% · output $1.32 → $0.141 -89%
• input $0.044 → $0.44 +891% · output $0.133 → $1.32 +890%
• input $0.042 → $0.044 +5% · output $0.169 → $0.133 -21%
• input $0.042 → $0.042 0% · output $0.176 → $0.169 -4%
• input $0.044 → $0.042 -5% · output $0.133 → $0.176 +32%

Providers on OpenRouter

25 endpoints serve this model. Uptime is OpenRouter's share of successful requests; prices per 1M tokens.