ST

Step 3.7 Flash

StepGGUF by stepfun-aiUpdated 2026-06-03
Download on Hugging Face
Parameters197B
Context—
Layers—
Architecture—
Smallest quant0.5 GB
Quick answer

Step 3.7 Flash needs about 105.6 GB of memory at Q3_K_M (a 93.8 GB file) with an 8K context window, which is more than a 48 GB GPU holds. This repository has 10 quants, from 0.5 GB to 394.0 GB (BF16).

Step 3.7 Flash – 198B sparse MoE vision‑language model

Step 3.7 Flash is a 198B‑parameter sparse Mixture‑of‑Experts model with a 196B language backbone and a 1.8B vision encoder. It supports a 256K context window and offers three reasoning levels to balance speed, cost, and depth.

Good at

  • High‑throughput multimodal inference
  • Native image understanding
  • Agentic workflow orchestration

Limits

  • Requires large unified memory for full context
  • No vision support in quantizations without mmproj
  • License is apache‑2.0

Best for

  • Parsing large financial reports
  • Running multi‑step search loops
  • Concurrent coding agents

Who it’s for. Developers building agentic workflows that combine perception, search, and reasoning. Suitable for intensive tasks such as parsing massive reports, multi‑step verification, or high‑throughput coding pipelines.

Quants & memory

QuantFileEst. memoryvs VRAMResultDownload
—0.5 GB 1.0 GB
Fits Get file →
Q8_03.7 GB 4.7 GB
Fits Get file →
BF167.0 GB 8.3 GB
Too big Get file →
IQ3_XXS75.8 GB (2 parts) 85.3 GB
Too big Get file →
Q3_K_M93.8 GB (3 parts) 105.6 GB
Too big Get file →
Q3_K_L102.5 GB (3 parts) 115.3 GB
Too big Get file →
IQ4_XS105.0 GB (3 parts) 118.1 GB
Too big Get file →
Q4_K_S111.5 GB (3 parts) 125.4 GB
Too big Get file →
Q8_0209.4 GB (5 parts) 235.0 GB
Too big Get file →
BF16394.0 GB (9 parts) 441.8 GB
Too big Get file →

For 128 GB unified memory, use Q4_K_S or IQ4_XS to run at full 256K context. If memory is limited, Q8_0 (3.71 GB) or BF16 (6.97 GB) provide smaller footprints but lower precision. Larger quants like Q3_K_L (102.5 GB) require significant memory and are best for high‑precision use cases.

What GPU do you need for Step 3.7 Flash?

These figures add the file size, an fp16 KV cache at an 8K context window and about 0.5 GB for the runtime. They are estimates, so leave a little headroom, and expect to use less context or a smaller quant if you are right at the limit.

  • 4 GB VRAM: the best quant that fits is the main file (0.5 GB file, about 1.0 GB in use).
  • 6 GB VRAM: the best quant that fits is Q8_0 (3.7 GB file, about 4.7 GB in use).
  • 8 GB VRAM: the best quant that fits is Q8_0 (3.7 GB file, about 4.7 GB in use).
  • 12 GB VRAM: the best quant that fits is BF16 (7.0 GB file, about 8.3 GB in use).
  • 16 GB VRAM: the best quant that fits is BF16 (7.0 GB file, about 8.3 GB in use).
  • 24 GB VRAM: the best quant that fits is BF16 (7.0 GB file, about 8.3 GB in use).
  • 32 GB VRAM: the best quant that fits is BF16 (7.0 GB file, about 8.3 GB in use).
  • 48 GB VRAM: the best quant that fits is BF16 (7.0 GB file, about 8.3 GB in use).

Which Step 3.7 Flash quant should you choose?

Q3_K_M (93.8 GB) is the middle option in this repository. A reasonable choice when a 4-bit file will not fit, mostly on larger models. Quality loss is noticeable but often acceptable.

If memory is tight, the smallest file is the first one in the table at 0.5 GB, with the quality trade-offs that come with fewer bits.

New to quant names? Read GGUF quantization explained.

Specifications

Architecture from the GGUF header

Parameters197B

Source

Quantized bystepfun-ai
FamilyStep
Licenseapache-2.0
Downloads (HF)6,198
Last updated2026-06-03

Frequently asked questions about Step 3.7 Flash

How much VRAM does Step 3.7 Flash need?

About 105.6 GB at Q3_K_M with an 8K context window. That is the 93.8 GB file, plus the KV cache and roughly 0.5 GB of runtime overhead. Smaller quants need less and longer context needs more.

Can I run Step 3.7 Flash on an 8 GB GPU?

Yes, with Q8_0: the 3.7 GB file needs about 4.7 GB at 8K context. Larger quants in this repository do not fit in 8 GB.

What is the file size of Step 3.7 Flash GGUF?

It depends on the quant: file is 0.5 GB, Q8_0 is 3.7 GB, BF16 is 7.0 GB, IQ3_XXS is 75.8 GB, Q3_K_M is 93.8 GB, Q3_K_L is 102.5 GB and 4 more.

Where can I download Step 3.7 Flash GGUF?

The files are on Hugging Face in the stepfun-ai/Step-3.7-Flash-GGUF repository. The download buttons on this page open the file on Hugging Face. ggufmodels.com does not host any model files.

What license does Step 3.7 Flash use?

Hugging Face lists the license as apache-2.0. Licenses can differ between the quantized files and the original model, so check the repository and the base model before commercial use.

Can I run Step 3.7 Flash on 8 GB VRAM?

No, the smallest quant requires 0.47 GB but still needs unified memory; 8 GB VRAM is insufficient for full context.

Does the model support vision tasks?

Yes, it includes a 1.8B vision encoder and mmproj projector for native image understanding.

What license applies to the model?

The model is released under the apache‑2.0 license.

Other GGUFs of Step-3.7-Flash

Similar models

Browse all →

More from stepfun-ai

All by stepfun-ai →

More Step models

All Step →

Other large models

All large models →