Step-3.7-Flash: hardware requirements
Step-3.7-Flash needs about 140 GB of memory at Q4 quantization — a 122 GB download plus room for context and overhead. That is beyond consumer hardware; use it through an API instead.
Parameters
201B
Tier
Server only
Licence
See model card
Released
2026-06-01
Will it run on your machine?
8GB RAMNot enough memory16GB RAMNot enough memory24GB RAMNot enough memory32GB RAMNot enough memory64GB RAMNot enough memory128GB RAMNot enough memory8GB VRAMNot enough memory12GB VRAMNot enough memory16GB VRAMNot enough memory24GB VRAMNot enough memory32GB VRAMNot enough memory
Download sizes by quantization
Lower quantization means a smaller file and less memory, at some cost to quality. Q4_K_M is the usual starting point. What is this?
| Quant | Download | RAM @ 4K | RAM @ 32K | Source |
|---|---|---|---|---|
| Q4_K_Mrecommended | 122 GB | 140 GB | 140 GB | measured |
| Q5_K_M | 146 GB | 167 GB | 167 GB | measured |
| Q6_K | 163 GB | 186 GB | 186 GB | measured |
| Q8_0 | 209 GB | 238 GB | 238 GB | measured |
| BF16 | 398 GB | 449 GB | 449 GB | measured |
“Measured” sizes are read from published GGUF files (unsloth/Step-3.7-Flash-GGUF). “Estimated” sizes are derived from the parameter count and are typically within a few percent. The RAM columns add the context cache and runtime overhead to the download size.
Model card on HuggingFace →API specs & pricing →Verified 2026-08-07Tracked for 12 months after release