Both current DeepSeek API models, deepseek-flash and deepseek-v4-pro, have a context window of 1M (one million) tokens, according to DeepSeek’s official Models & Pricing page.
As of November 9, 2026, deepseek-flash is served by DeepSeek-V4.1-Flash-1017 and deepseek-v4-pro by DeepSeek-V4-Pro-0813. DeepSeek lists a maximum output of 384K tokens, and both models support thinking and non-thinking modes, JSON output and tool calls. The legacy names deepseek-v4-flash and deepseek-v4-flash-vision-exp still work, but their requests now run on the V4.1 Flash model. Pages that quote smaller limits usually describe earlier DeepSeek models rather than the current V4-series API.
Long prompts are billed per token: deepseek-flash costs $0.30 per million cache-miss input tokens and $1.20 per million output tokens at peak rates, and off-peak rates are half that.
Related: DeepSeek API pricing: how much does the API cost?, What are the DeepSeek API rate limits? and DeepSeek Harness guide
Source: DeepSeek API Docs, Models & Pricing
Image: DeepSeek logo icon, CC0 public domain dedication, via Wikimedia Commons. Placed on a white background and resized. DeepSeek is a trademark of its owner. FreeAnswers is not affiliated with or endorsed by DeepSeek.