Skip to content

Step 3.5 Flash on DGX Spark: VRAM, speed and fit

Step 3.5 Flash at 3-bit runs on the DGX Spark 128 GB with 56K of context at about 37 tokens a second.

Text generation · Chat · Ollama / LM Studio

Runs comfortably · on DGX Spark 128 GB · 3-bit (Q3_K_M) · Ollama / LM Studio · Chat

GPU memory
121 GB
of 128 GB
Context
56K
auto
Speed
37 tok/s
6.8 when full
First reply
2.0 s
2K prompt

Where the memory goes

Model
90.8 GB
Context
21.6 GB
Overhead
636 MB
OS
7.7 GB
Free
7.4 GB
Memory
121 GB of 128 GB
Context
56K tokens · auto
Speed
37 tok/s · 6.8 when full
First reply
2.0 s for a 2K prompt
Running cost
$3.15/mo at home · $287/mo on a cloud 4× A40

Validation: Validated

What Step 3.5 Flash on DGX Spark was checked against
CheckTargetResultStatus
Fits in GPU memory120 GB usable113 GB neededPass
Headroom5% free7.4 GB free (6%)Pass
Context for chat16K or more56KPass
Writing speed15 tok/s or more37 tok/sPass
First replyunder 3 s, 2K prompt2.0 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
3-bit (Q3_K_M)
Context
56K, auto
Chats at once
1
Use
Chat
Runs in
Ollama / LM Studio
Hardware
DGX Spark 128 GB
GPU can use
120 GB of 128 GB

Worth trying

  • 8-bit cache: Room for 112K context

Running cost

At home
$3.15/mo
144 W at the wall
Renting instead
$287/mo
4× A40 at $2.36/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 Spark run?

Other models that run on DGX Spark

Questions

Can Step 3.5 Flash run on DGX Spark?

Step 3.5 Flash at 3-bit runs on the DGX Spark 128 GB with 56K of context at about 37 tokens a second.

Report VR-0LLQ4NW, calculated Nov 22, 2026 from published specs, not benchmarked.