Woofun AI reports that Kimi Code has officially transitioned the K2.7 Code HighSpeed mode from beta to a permanent feature. This version utilizes the same underlying model as the standard offering but achieves output rates of approximately 180 Tokens/s in typical scenarios, peaking at 260 Tokens/s for short contexts.

Users with Allegretto tier or higher subscriptions can immediately access the HighSpeed mode through Kimi Code CLI or compatible Coding Plan tools without additional application. Activation requires CLI users to input '/model' and select 'kimi-for-coding-highspeed'. The HighSpeed variant consumes three times the usage quota of the regular version, positioning it specifically for coding tasks where response latency is the primary constraint.