Unverified50% confidenceFactExact time
Anthropic published a research paper showing that a relatively small 9-billion-parameter model successfully completed AI inference setup on 8 of the 9 hardware types tested, and successfully ran a 26-billion-parameter model on 15 cheap Android TV boxes.
1
Sources
50%
Confidence
Long-term
Relevance
10/14/2026
First Seen
Sources
Related Entities
Related Claims
UnverifiedA 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.74% similarUnverifiedOn-device AI can typically only run small quantized models with parameter counts in the range of 1 billion to 7 billion.71% similarUnverified苹果的端侧模型号称有200亿参数，采用MoE式部分激活机制，每次推理只激活10亿到40亿参数70% similarUnverified阿里千问新一代模型参数达2.4万亿，是该系列首个突破万亿的多模态大模型，代码生成准确率据称达到94.7%，较上一代提升约11个百分点70% similarUnverified开发团队用分布式拆层方法测试过高达1万亿参数的模型70% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/1030341API
curl https://kongchang.com/api/v1/knowledge/claims/1030341MCP
get_claim(id=1030341)