Loading…
Get the best builds by email. No account needed.
Live drop
TencentDropped Nov 16, 2026
1M-context multimodal model at $0.07/0.28 per million tokens for daily productivity.
About this release
Hy4 flash is Tencent's lowest-cost Hy4 variant: 224B total parameters (12B active) with native multimodal input, 1M-token context, and optimized inference exceeding 150 tokens per second. It targets document editing, slide creation, research, and visual tasks at roughly 1/20th the per-task cost of the full Hy4 flagship.
Read the official Tencent launch pageFirst builds incoming — this lane fills in automatically as people start building with Hy4 flash. Know one? Submit the URL.