DeepSeek-V4-Flash-0731
The official release of DeepSeek-V4-Flash with enhanced agentic capabilities.
A specific language-model release identified in the supplied evidence as DeepSeek V4 Flash / Flash-0731, used for coding-agent and long-context workloads.
Pricing
Official current Flash rates per 1M tokens: peak cache-miss input $0.30, output $1.20, cache-hit input $0.006. Off-peak rates are half: $0.15 input cache miss, $0.60 output, $0.003 cache hit. Rates apply to currently served DeepSeek-V4.1-Flash-1017; exact 0731 direct-serving rate is not stated.
DeepSeek’s official docs state that legacy deepseek-v4-flash requests are retired and served by DeepSeek-V4.1-Flash-1017 at the current Flash rate. They do not explicitly identify the exact 0731 checkpoint as a current direct API model, so these are current DeepSeek Flash API rates, not a separately verified serving price for the unchanged 0731 checkpoint.