Skip to content

Ornith 1.0 397B on 8× H200 NVL: VRAM, speed and fit

Ornith 1.0 397B at 8-bit runs on the 8× H200 NVL with 256K of context at about 108 tokens a second.

Text generation · Chat · Ollama / LM Studio

Runs comfortably · on 8× H200 NVL · 8-bit (Q8_0) · Ollama / LM Studio · Chat

GPU memory
407 GB
of 1,128 GB
Context
256K
auto
Speed
108 tok/s
86 when full
First reply
0.1 s
2K prompt

Where the memory goes

Model
393 GB
Context
7.7 GB
Overhead
6.3 GB
OS
0.0 GB
Free
721 GB
Memory
407 GB of 1,128 GB
Context
256K tokens · auto
Speed
108 tok/s · 86 when full
First reply
0.1 s for a 2K prompt
Running cost
$65/mo at home · $1,162/mo on a cloud 4× MI300X

Validation: Validated

What Ornith 1.0 397B on 8× H200 NVL was checked against
CheckTargetResultStatus
Fits in GPU memory1,128 GB usable407 GB neededPass
Headroom5% free721 GB free (64%)Pass
Context for chat16K or more256KPass
Writing speed15 tok/s or more108 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
Ornith 1.0 397B
Parameters
397B, 17.8B active
Precision
8-bit (Q8_0)
Context
256K, auto
Chats at once
1
Use
Chat
Runs in
Ollama / LM Studio
Hardware
8× H200 NVL
GPU can use
1,128 GB of 1,128 GB

Running cost

At home
$65/mo
2960 W at the wall
Renting instead
$1,162/mo
4× MI300X at $9.56/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 unsloth/Ornith-1.0-397B-GGUF
vLLM
vllm serve ornith-ai/Ornith-1.0-397B

Ornith 1.0 397B on similar machines

All machines: Ornith 1.0 397B VRAM requirements. Everything that runs here: what can the 8× H200 NVL run?

Other models that run on 8× H200 NVL

Questions

Can Ornith 1.0 397B run on 8× H200 NVL?

Ornith 1.0 397B at 8-bit runs on the 8× H200 NVL with 256K of context at about 108 tokens a second.

Report VR-05J2ONE, calculated Nov 22, 2026 from published specs, not benchmarked.