Hy4 Flash
A low-cost large language model introduced by Tencent that inherits the full feature set of Hy4. It supports reasoning, tool calling, structured output, prompt caching and image input, offers reasoning capability that can be disabled, and has a context window of 1 million tokens.
Timeline (last 90 days)
Tencent's Hy4 Flash model is now available on AI Gateway.
Hy4 Flash is offered on AI Gateway at a 50% introductory discount price, covering input, cached input and output tokens, through November 29, 2026.
Hy4 Flash is a new lower-cost member of the Hy4 family, retaining the full feature set of Hy4 at a per-token price of about one-tenth of Hy4.
The features Hy4 Flash retains include reasoning capability that preserves the thinking process across turns, tool calling, structured outputs, prompt caching and image input, while maintaining a 1 million token context window.
The official recommendation is that when using Hy4 Flash in coding agents such as Claude Code, Codex and Cursor, install the latest version of the Vercel CLI and run `npx vercel ai-gateway setup`, then select `tencent/hy4-flash`.
All Facts (5)
Hy4 Flash is offered on AI Gateway at a 50% introductory discount price, covering input, cached input and output tokens, through November 29, 2026.
50%UnverifiedHy4 Flash is a new lower-cost member of the Hy4 family, retaining the full feature set of Hy4 at a per-token price of about one-tenth of Hy4.
50%UnverifiedThe features Hy4 Flash retains include reasoning capability that preserves the thinking process across turns, tool calling, structured outputs, prompt caching and image input, while maintaining a 1 million token context window.
50%UnverifiedThe official recommendation is that when using Hy4 Flash in coding agents such as Claude Code, Codex and Cursor, install the latest version of the Vercel CLI and run `npx vercel ai-gateway setup`, then select `tencent/hy4-flash`.
50%UnverifiedTencent's Hy4 Flash model is now available on AI Gateway.
50%