Unverified50% confidenceFactExact time
目前能在手机端流畅运行的模型参数量通常在30亿至70亿级别，与云端数千亿参数的旗舰模型存在数量级差距
1
Sources
50%
Confidence
Medium-term (~90 days)
Relevance
8/8/2026
First Seen
Valid until: 11/6/2026
Sources
苹果AI战略真的失误了吗？深度解析行业质疑与苹果的隐藏布局
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.70% similarUnverified根据DeepMind在2022年提出的Chinchilla缩放定律，一个700亿参数的模型需要约1.4万亿token才能达到计算最优67% similarUnverifiedThe GPU resources required for a single inference of a hundred-billion-parameter flagship model may be 10 to 20 times higher than those of a ten-billion-parameter model.65% similarUnverifiedKimi 2.5 computes only approximately 30 billion parameters per forward pass, despite having a total parameter count of 1 trillion.65% 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.65% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/707102API
curl https://kongchang.com/api/v1/knowledge/claims/707102MCP
get_claim(id=707102)