Back to all models
deepseek
deepseekDeepSeek: 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....
deepseek/deepseek-v4-flash-0731
Context Size
1,310.72K
Input Price
9,000 pts/M
Output Price
18,000 pts/M
Architecture
Text
Supported Parameters
frequency_penaltyinclude_reasoninglogit_biaslogprobsmax_tokensmin_pparallel_tool_callspresence_penaltyreasoningreasoning_effortrepetition_penaltyresponse_formatseedstopstructured_outputstemperaturetool_choicetoolstop_atop_ktop_logprobstop_p
Details
TokenizerDeepSeek
Max Completion943,718 tokens
Provider Context1,048.576K tokens
ModeratedNo
CreatedJul 31, 2026