AI modelsNovember 16, 2026
-16
āļø
Tencent's Hy4 Flash Slashes AI Costs
Tencent launches Hy4 flash with 12 billion active parameters, a 1-million-token context, and very low pricing.
#Tencent#LLM#MoE#multimodal#pricing

š„ What happened
Tencent launched Hy4 Flash: 224B total parameters but only 12B active. It costs just $0.07 per million input tokens ā a fraction of rivals.
š” Why it matters
Hy4 Flash is 13x cheaper per task than Tencent's own Hy4 Preview and 8x cheaper than DeepSeek-V4.1-Flash, per internal benchmarks. With 1M context and 150 tokens/sec, it targets everyday productivity work.
ā” Our take
Tencent is no longer optimizing for benchmarks but for cost per completed task. Anyone still buying expensive models for routine work is burning money.
The title, summary and analysis of this item were produced automatically by an AI system and have not been editorially reviewed. They may contain errors, bias or omissions ā when in doubt, read the linked original source.
Deep Dives & Similar Intelligence

Alibaba Strikes Back with Qwen4-Max
AI modelsNovember 16, 2026
Alibaba open-sources its code review agent
Tools & ProjectsNovember 13, 2026

165M Model Trained for $4,200
ResearchNovember 15, 2026

Microsoft VP ships GPU task manager
Tools & ProjectsNovember 12, 2026

EchoMoE Slashes MoE Communication by 58%
ResearchNovember 17, 2026