Hy3, the new 295B MoE model from @TencentHunyuan, is now free in Nous Portal for the next two weeks! It is focused on cost-effective agentic use, and particularly strong on coding, tool-calling reliability, reasoning, and 256K long-context tasks. https://t.co/2uiUhfas0P
Nous Research Adds Tencent Hy3 for Reliable and Cost-Effective Agentic Workflows
Nous ResearchNous Research integrated Tencent's new Hy3 model into its Portal, offering free access for the next two weeks. The model is optimized for agentic workflows, prioritizing tool-calling reliability and reasoning for complex, multi-step tasks.
Hy3 model to its unified subscription platform. Hy3 uses a Mixture-of-Experts (MoE) architecture (an architecture activating only a fraction of parameters per request) to balance high-scale reasoning with inference efficiency.This integration targets the growing demand for agentic models—systems that autonomously plan and use tools—that prioritize tool-calling reliability and reasoning for multi-step tasks. By focusing on "intelligence density," the model aims to compete with frontier systems while maintaining cost-efficiency.
You can access Hy3 for free on the Nous Portal for the next 14 days. The platform uses Nous Portal's credit management to provide a single API key for over 300 models, including Step 3.7 Flash and credits for Hermes Agent v0.18.0 tool usage.
Still wondering? A few quick answers below.
Every HeadsUpAI update is written based on its original source and reviewed before it's published. Read our editorial standards →



