Hy4 Flash
Tencent has introduced Hy4 flash, its most cost-effective AI model to date, featuring 224B total parameters and 12B active parameters per token. The natively multimodal model supports a 1M-token context window while significantly lowering API pricing and achieving high decoding speeds past 150 tokens per second. Hy4 flash is now available across several Tencent services and platforms, with plans to release it as an open-weight model. Open article →
- Hy4 Flash now available on AI Gateway for 50% off
- Tencent Hy4 flash Benchmark Results for Data Analysis
- Tencent launches Hy4 flash, cutting AI task costs by 95% in internal tests
- BREAKING: Tencent just shipped Hy4 Flash. Haiku-level quality on everyday work at roughly 1/18th...
- Tencent didn't waste any time. Just three hours after Alibaba announced Qwen4-Max, Tencent came o...
- Tencent just released the most interesting home-lab model of the month. Hy4 flash is 224B total,...
- The cost column in Tencent's chart is the part people will skip, and it's the part that matters. ...
- Hy4 Flash is a real efficiency flex‼️ 224B total/only 12B active (+ a 64B Engram lookup module),...
- Meet Hy4 flash. Skip the pileup. Let's take the Hyway ◆ Best value Hy yet: $0.07 in, $0.281 out p...
- Hy4 flash - the CHEAPEST & FASTEST model USEFUL for MOST everyday tasks! RMB 0.2 / USD 0.0...
- This is Hy4 Flash btw 🔥 - 224B total parameters and 12B active - Should fit 2x DGX Sparks! Open...
- OK it's not top tier but it's VERY cheap it's equal to Hy4-Preview but it's a fraction of the cos...
- HY4 Flash API is out. A few takeaways: - HY4 family is working with DS Flash, instead of trying ...
- ⚡Meet Hy4 flash, the cheapest model for everyday productivity. Matches Claude Haiku 5.5 (high) on...