@thsottiaux Token optimization. Some tasks will burn insane amounts for seemingly no reason. Would be great for dynamic think / intelligence ie tasks are automatically routed for token efficiency and quality. Like I’ll spend where it counts but want codex to select the best think
The new Hy3-Preview model from @TencentHunyuan is live for free on OpenRouter!
It’s a 295B MoE model (21B active) with controllable reasoning effort.
A cost-effective, practical model that performs strongly in coding agents and delivers comprehensive general capabilities.