Unverified50% confidenceFactExact time
Moonshot AI's Kimi K3 has 2.8 trillion parameters, with model files exceeding 1,500 GB.
1
Sources
50%
Confidence
Medium-term (~90 days)
Relevance
9/22/2026
First Seen
Valid until: 12/21/2026
Sources
Related Entities
Related Claims
UnverifiedRunning Kimi K3 requires approximately 700GB of VRAM, and the community estimates it needs an H200 cluster worth $3 million or multiple data-center-grade devices.75% similarVerifiedKimi K3 supports an ultra-long context window of 1 million tokens.73% similarUnverifiedA developer ran the 2.8 trillion-parameter Kimi K3 model on a single MacBook Pro at approximately 1 token/second.70% similarUnverifiedKimi K2在FP16精度下仅模型权重就需要数百GB显存，INT4量化可将显存需求压缩至约四分之一68% similarVerifiedWhen models with hundreds of billions of parameters process a 128K context window, KV Cache can occupy tens of gigabytes of VRAM.67% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/941006API
curl https://kongchang.com/api/v1/knowledge/claims/941006MCP
get_claim(id=941006)