Unverified50% confidenceBenchmarkExact time
在M4 Max上进行FP16精度测试，hy3d shape（small）任务运行时间20.9秒，峰值内存约5.6GB
1
Sources
50%
Confidence
Long-term
Relevance
7/13/2026
First Seen
Sources
Apple Silicon本地图生3D：20秒生成、2GB内存，Modelr开源应用实测
Related Claims
UnverifiedWhen running at 1-bit precision on an M2 Max equipped with 96GB of VRAM, generation speed for simple tasks is consistently around 28–30 tokens/second73% similarUnverifiedThe Gemma 4 26B MOE version achieves an actual inference speed of 60 tokens per second on a GPU with 24GB of VRAM.70% similarUnverifiedA tester ran a 20.7B dense model on a dual-4090 (48 GB VRAM) workstation and achieved 27.6 tokens per second, only slightly faster than Dwarf Star, but with over 10 times fewer parameters.69% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/502820API
curl https://kongchang.com/api/v1/knowledge/claims/502820MCP
get_claim(id=502820)