The story
Tencent has launched Hy4 Flash, a cost-focused version of its Hy4 model, with 224 billion total parameters and 12 billion active in its mixture-of-experts design. The model offers a one-million-token context window and is priced at $0.07 per million input tokens and $0.281 per million output tokens. TechNode reports that on Tencent's internal WorkIndex benchmark, Hy4 Flash costs one-twentieth as much per task as the full Hy4 — a vendor claim that has not been independently tested. The weights have not yet been released, though a listing on OpenRouter has been promised.Latest development
cooling, formula 66.0, stored 46.0