← EVENT LIBRARY / Foundation Models

Moonshot launches Kimi K3.1 Pro before releasing its weights

Moonshot launched Kimi K3.1 Pro and a Highspeed tier across Kimi, Kimi Work, Kimi Code and its API. The model has a one-million-token context window, while Highspeed is advertised at two to three times the serving speed. Full weights are promised by November 4 but are not yet available; benchmark claims are publisher-reported.

Availability: Kimi, Kimi Work, Kimi Code and Kimi API available now; full weights promised by November 4

What this means for you

For coding and long-context teams, the callable API enables immediate evaluation and the promised weights could materially widen deployment options. Availability and performance must be kept separate: the API is live, but the open weights and independent benchmark runs are not.

Evidence & related material

Technical details & scores
Model Class 3T-class model; exact architecture and active-parameter count not stated on the release page
Context Window 1 million tokens
Availability Kimi, Kimi Work, Kimi Code and Kimi API available now; full weights promised by November 4
Speed Claim Highspeed advertised at 2-3x K3.1 Pro speed
Evaluation Boundary benchmark results are publisher-reported; Kimi Code Bench 3.0 is internal
Pricing Boundary no API price was verified from the release page
Time Basis the official page establishes October 23 at date precision; Hacker News submission at 2026-10-23T16:59:08Z is only an external upper bound, not the official publication time
Technical impact 4
Engineering value 4
Adoption signal 4
Maturity 4
Verification cost 3
Risk 3