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.
Details
| Field | Value |
|---|---|
| id | deepseek/deepseek-v4-flash-0731 |
| family | deepseek-flash |
| modality | text->text |
| context_length | 1048576 |
| max_output_tokens | 943718 |
| knowledge_cutoff | 2025-05 |
| release_date | 2026-07-31 |
| is_open_weights | yes |
| supports_tool_call | yes |
| supports_reasoning | yes |
| supports_structured_output | yes |
| supports_attachment | no |
| input_modalities | text |
| output_modalities | text |