Verified65% confidenceBenchmarkExact time
MiniMax H3 still requires close to 24 GB of VRAM under NF4 quantization.
3
Sources
65%
Confidence
Long-term
Relevance
8/26/2026
First Seen
Sources
Related Entities
Related Claims
UnverifiedThe BF16 full model of MiniMax H3 is over 50 GB in size and requires close to 60 GB of VRAM to load completely.79% similarUnverifiedMiniMax H3's weights total 67 GB, comprising a 34 GB DiT backbone, a 27 GB Qwen3-VL encoder, plus VAE77% similarVerifiedModels with 32B parameters can typically be compressed to approximately 16–20 GB of VRAM usage after quantization, and can be run on a single NVIDIA RTX 4090 (24 GB VRAM).75% similarUnverifiedRunning the MiniMax H3 base model is recommended with a GPU with 24GB or more VRAM, such as an NVIDIA RTX 4090 or A100.74% similarUnverifiedThe MiniMax H3 standard workflow requires a minimum of 16GB VRAM to run, while 6GB graphics cards can also participate in partial workflows through optimizations such as FP8, INT4 quantization, and chunked inference Tiling.71% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/805289API
curl https://kongchang.com/api/v1/knowledge/claims/805289MCP
get_claim(id=805289)