Moonshot AI has given users direct control over how hard its new flagship model thinks, rolling out three selectable reasoning effort levels — Standard, High and Max — for Kimi K3 on the Kimi web app and mobile client. The update, arriving within a day of the model's launch, lets people trade depth of reasoning against speed and cost rather than being locked into a single always-maximal mode.
Continue reading
The rest of this article is for AI News Blitz readers. Choose an option below to keep reading.
Already purchased? Sign in✓ Signed in — this article isn’t included in your current plan.