Tencent has launched Hy3, a coding-oriented AI model priced at $0.14 per million input tokens, and the number is the story. In a market where frontier coding models charge multiples of that, a sub-fifteen-cent rate fundamentally changes the economics of running agent fleets at scale.
Price is a feature, especially for agents. A coding agent doesn’t make one call — it makes hundreds of tool calls, retries, and context reloads per task. When input tokens are nearly free, the whole calculus of “should I re-read the whole repo or just the diff” tilts toward thoroughness. Cheap input means agents can be more context-hungry without blowing the budget.
The strategic angle is familiar from the API price wars: a low entry price becomes the default fallback router for cost-sensitive steps, even if it isn’t the absolute best at the hardest problems. Orchestrators will happily slot Hy3 in for the boring 80% and reserve the pricey flagship for the gnarly 20%.
For builders, Hy3 is a reminder that the winning model isn’t always the smartest — sometimes it’s the one you can afford to call a thousand times an hour.
Smart developers shop around. Find the best AI tool deals at aiFiesta.