Skip to content

Ornith 1.5 397B on DGX Station GB300: VRAM, speed and fit

Ornith 1.5 397B at 4-bit runs on the DGX Station GB300 with 256K of context at about 151 tokens a second.

Text generation · Chat · Ollama / LM Studio

Runs comfortably · on DGX Station GB300 · 4-bit (Q4_K_M) · Ollama / LM Studio · Chat

GPU memory
236 GB
of 252 GB
Context
256K
auto
Speed
151 tok/s
121 when full
First reply
0.4 s
2K prompt

Where the memory goes

Model
228 GB
Context
7.7 GB
Overhead
1.0 GB
OS
0.0 GB
Free
15.5 GB
In RAM
0.0 GB
Memory
236 GB of 252 GB
Context
256K tokens · auto
Speed
151 tok/s · 121 when full
First reply
0.4 s for a 2K prompt
Running cost
$23/mo at home · $574/mo on a cloud 8× A40

Validation: Validated

What Ornith 1.5 397B on DGX Station GB300 was checked against
CheckTargetResultStatus
Fits in GPU memory252 GB usable236 GB neededPass
Headroom5% free15.5 GB free (6%)Pass
Context for chat16K or more256KPass
Writing speed15 tok/s or more151 tok/sPass
First replyunder 3 s, 2K prompt0.4 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
Ornith 1.5 397B
Parameters
403B, 24.4B active
Precision
4-bit (Q4_K_M)
Context
256K, 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 zephel01/Ornith-1.5-397B-IQ2_XXS-32G128G-GGUF
vLLM
vllm serve ornith-ai/Ornith-1.5-397B

Ornith 1.5 397B on similar machines

All machines: Ornith 1.5 397B VRAM requirements. Everything that runs here: what can the DGX Station GB300 run?

Other models that run on DGX Station GB300

Questions

Can Ornith 1.5 397B run on DGX Station GB300?

Ornith 1.5 397B at 4-bit runs on the DGX Station GB300 with 256K of context at about 151 tokens a second.

Report VR-1W45F67, calculated Nov 22, 2026 from published specs, not benchmarked.