OpenRouter said on November 16, 2026 that Tencent Hunyuan's Hy4 flash is live on its platform. According to the provider's announcement on X, the model is a faster and cheaper version of Hy4 while retaining a one-million-token context window and image input.
OpenRouter is offering the model at a 50% discount through November 30: $0.035 per million input tokens and $0.14 per million output tokens.
Tencent released Hy4 flash on the same day as the first Flash-tier model in the series. Independent coverage of Tencent's announcement reports 224 billion total parameters, of which 12 billion are active per inference, and multimodal input. Tencent positions the model for everyday work such as research, document processing and presentations, while recommending the larger Hy4 for more complex tasks.