Unverified50% confidenceFactExact time
The Kimi 3.1 Pro specifications circulating in the community show that it uses a 104BA-2.8T architecture and is an open-weight model.
1
Sources
50%
Confidence
Short-term (~14 days)
Relevance
10/24/2026
First Seen
Valid until: 11/7/2026
Sources
Related Entities
Related Claims
UnverifiedKimi K3.1 Pro and K3.1 Pro Highspeed are two presentations of the same underlying model in terms of inference configuration.70% similarUnverifiedKimi K3.1 uses a MoE architecture, with about 2.8 trillion total parameters and 104 billion activated parameters.69% similarUnverifiedKimi has released an all-new generation model, Kimi K3.69% similarVerifiedKimi K3 almost certainly adopts a Mixture of Experts (MoE) architecture.66% similarUnverifiedvLLM v0.27.0 provides complete full-stack support for the Kimi K3 model, with all related components landing at once in the same version.64% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/1092639API
curl https://kongchang.com/api/v1/knowledge/claims/1092639MCP
get_claim(id=1092639)