Skip to content

Step 3.5 Flash on 8× RTX PRO 6000 Blackwell: VRAM, speed and fit

Step 3.5 Flash at 8-bit runs on the 8× RTX PRO 6000 Blackwell with 256K of context at about 94 tokens a second.

Text generation · Chat · Ollama / LM Studio

Runs comfortably · on 8× RTX PRO 6000 Blackwell · 8-bit (Q8_0) · Ollama / LM Studio · Chat

GPU memory
299 GB
of 768 GB
Context
256K
auto
Speed
94 tok/s
11 when full
First reply
0.1 s
2K prompt

Where the memory goes

Model
197 GB
Context
96.6 GB
Overhead
5.3 GB
OS
0.0 GB
Free
469 GB
Memory
299 GB of 768 GB
Context
256K tokens · auto
Speed
94 tok/s · 11 when full
First reply
0.1 s for a 2K prompt
Running cost
$65/mo at home · $574/mo on a cloud 8× A40

Validation: Validated

What Step 3.5 Flash on 8× RTX PRO 6000 Blackwell was checked against
CheckTargetResultStatus
Fits in GPU memory768 GB usable299 GB neededPass
Headroom5% free469 GB free (61%)Pass
Context for chat16K or more256KPass
Writing speed15 tok/s or more94 tok/sPass
First replyunder 3 s, 2K prompt0.1 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
256K, auto
Chats at once
1
Use
Chat
Runs in
Ollama / LM Studio
Hardware
8× RTX PRO 6000 Blackwell
GPU can use
768 GB of 768 GB

Running cost

At home
$65/mo
2960 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 8× RTX PRO 6000 Blackwell run?

Other models that run on 8× RTX PRO 6000 Blackwell

Questions

Can Step 3.5 Flash run on 8× RTX PRO 6000 Blackwell?

Step 3.5 Flash at 8-bit runs on the 8× RTX PRO 6000 Blackwell with 256K of context at about 94 tokens a second.

Report VR-09565L0, calculated Nov 22, 2026 from published specs, not benchmarked.