All models

Moonshot

Kimi K2.6 Fast

Kimi K2.6 on a low-latency route — the first token comes back faster than the default, in exchange for lower per-route capacity.

1T MoE (~32B active)Modified MIT
Contact sales

Capacity for Kimi K2.6 Fast isn't provisioned in our public fleet yet. For early access, benchmarks, or a custom deployment, talk to sales — either via the /enterprise form or by emailing cogito@decart.ai.

Use cases

Coding agentsLatency-sensitive tool useLong-horizon reasoning