Expired50% confidenceFactExact time
Apple's on-device model reportedly has 20 billion parameters and uses an MoE-style partial activation mechanism, activating only 1 billion to 4 billion parameters per inference.
1
Sources
50%
Confidence
Medium-term (~90 days)
Relevance
7/21/2026
First Seen
This claim has expired. It is kept for the record and no longer describes the current state.
Valid until: 10/19/2026(expired)
Sources
Related Claims
UnverifiedSome iPhones are equipped with a local model with 20 billion parameters that uses a sparse architecture, with only 1 billion to 4 billion parameters active at any given moment.82% similarUnverifiedA model with 100 billion parameters requires approximately 80 GB of VRAM to process a single inference request, while a model with tens of billions of parameters requires only about 20 GB.73% similarUnverifiedAnalysis speculates that the model is an MoE model with 2 to 4 trillion parameters, activating only approximately 150 billion parameters per token.69% similarUnverifiedBefore outputting a final answer, a reasoning model's internal chain of thought may consume 3,000 to 20,000 tokens, far exceeding the approximately 500 tokens of an ordinary model.68% similarUnverified参数量仅10亿至70亿的小语言模型在特定任务上可以逼近参数量百倍于它的大模型66% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/584665API
curl https://kongchang.com/api/v1/knowledge/claims/584665MCP
get_claim(id=584665)