Skip to content

Step 3.5 Flash on DGX Station GB300: VRAM, speed and fit

Step 3.5 Flash at 8-bit runs on the DGX Station GB300 with 104K of context at about 212 tokens a second.

Text generation · Chat · Ollama / LM Studio

Runs comfortably · on DGX Station GB300 · 8-bit (Q8_0) · Ollama / LM Studio · Chat

GPU memory
238 GB
of 252 GB
Context
104K
auto
Speed
212 tok/s
75 when full
First reply
0.2 s
2K prompt

Where the memory goes

Model
197 GB
Context
39.6 GB
Overhead
841 MB
OS
0.0 GB
Free
14.3 GB
In RAM
0.0 GB
Memory
238 GB of 252 GB
Context
104K tokens · auto
Speed
212 tok/s · 75 when full
First reply
0.2 s for a 2K prompt
Running cost
$23/mo at home · $574/mo on a cloud 8× A40

Validation: Validated

What Step 3.5 Flash on DGX Station GB300 was checked against
CheckTargetResultStatus
Fits in GPU memory252 GB usable238 GB neededPass
Headroom5% free14.3 GB free (6%)Pass
Context for chat16K or more104KPass
Writing speed15 tok/s or more212 tok/sPass
First replyunder 3 s, 2K prompt0.2 sPass

Targets: a reply you can read along with, a first reply the use case tolerates, and 5% of memory left free. Calculated from published specs, not benchmarked.

The setup

Model
Step 3.5 Flash
Parameters
199B, 9.4B active
Precision
8-bit (Q8_0)
Context
104K, auto
Chats at once
1
Use
Chat
Runs in
Ollama / LM Studio
Hardware
DGX Station GB300
GPU can use
252 GB of 252 GB
Spill space
480 GB LPDDR5X on the Grace CPU

Running cost

At home
$23/mo
1040 W at the wall
Renting instead
$574/mo
8× A40 at $4.72/h

4 h a day, power at $0.18 per kWh, cloud at RunPod on-demand prices as of Nov 22, 2026.

How to run it

llama.cpp
llama-server -hf ggml-org/Step-3.5-Flash-GGUF
vLLM
vllm serve stepfun-ai/Step-3.5-Flash
MLX (Mac)
mlx_lm.chat --model mlx-community/Step-3.5-Flash-4bit

Step 3.5 Flash on similar machines

All machines: Step 3.5 Flash VRAM requirements. Everything that runs here: what can the DGX Station GB300 run?

Other models that run on DGX Station GB300

Questions

Can Step 3.5 Flash run on DGX Station GB300?

Step 3.5 Flash at 8-bit runs on the DGX Station GB300 with 104K of context at about 212 tokens a second.

Report VR-0CLGMA5, calculated Nov 22, 2026 from published specs, not benchmarked.