Unverified50% confidenceBenchmarkExact time
MiniMax H3 在 NF4 量化下仍需接近 24GB 显存
1
Sources
50%
Confidence
Long-term
Relevance
8/26/2026
First Seen
Sources
Related Entities
Related Claims
Verified32B参数规模的模型在量化后通常可以压缩到约16-20GB显存占用，一张NVIDIA RTX 4090（24GB显存）就能运行75% similarUnverifiedMiniMax H3标准工作流最低16G显存即可运行，6G显卡通过FP8、INT4量化和分块推理Tiling等优化也能参与部分流程71% similarUnverifiedINT4 量化模型有望在 8GB 乃至更低显存的显卡上完成原本需要 16GB 或 24GB 显存的模型推理71% similarUnverifiedLlama 3 70B量化版（Q4精度）仍需约35GB内存71% similarUnverified目前单颗HBM3e堆叠最大约36GB70% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/805289API
curl https://kongchang.com/api/v1/knowledge/claims/805289MCP
get_claim(id=805289)