DeepSeek’s pricing page, checked 17 November 2026, says the current Flash API model name is deepseek-flash and the version is DeepSeek-V4.1-Flash-1017. It lists a 1M context window, a 384K maximum output, vision, tool calls, and thinking mode (on by default, with a documented way to switch). Peak hours are 01:00–04:00 and 06:00–10:00 UTC, Monday through Friday, excluding Chinese public holidays; off-peak rates are half of peak.
Hy4 flash, on Tencent TokenHub, also has a 1M context window, but the API guide caps input at 960k tokens and output at 64k tokens. List prices do not split peak and off-peak. Tencent’s launch post compared average WorkIndex task cost with DeepSeek-V4.1-Flash-1017 (Max tier, peak-hour pricing) and said Hy4 flash was one-eighth of that cost. We quote that sentence as a vendor claim only.