Skip to content

Ornith 1.5 397B on M5 Ultra 512 GB: VRAM, speed and fit

Ornith 1.5 397B at 6-bit runs on the M5 Ultra 512 GB with 256K of context at about 29 tokens a second.

Text generation · Chat · Ollama / LM Studio

Runs comfortably · on M5 Ultra 512 GB · 6-bit (Q6_K) · Ollama / LM Studio · Chat

GPU memory
445 GB
of 512 GB
Context
256K
auto
Speed
29 tok/s
22 when full
First reply
4.6 s
2K prompt

Where the memory goes

Model
308 GB
Context
7.7 GB
Overhead
869 MB
OS
128 GB
Free
67.4 GB
Memory
445 GB of 512 GB
Context
256K tokens · auto
Speed
29 tok/s · 22 when full
First reply
4.6 s for a 2K prompt
Running cost
$5.06/mo at home · $574/mo on a cloud 8× A40

Validation: Validated with conditions

What Ornith 1.5 397B on M5 Ultra 512 GB was checked against
CheckTargetResultStatus
Fits in GPU memory384 GB usable317 GB neededPass
Headroom5% free67.4 GB free (18%)Pass
Context for chat16K or more256KPass
Writing speed15 tok/s or more29 tok/sPass
First replyunder 3 s, 2K prompt4.6 sCheck

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
6-bit (Q6_K)
Context
256K, auto
Chats at once
1
Use
Chat
Runs in
Ollama / LM Studio
Hardware
M5 Ultra 512 GB
GPU can use
384 GB of 512 GB

Running cost

At home
$5.06/mo
231 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 M5 Ultra 512 GB run?

Other models that run on M5 Ultra 512 GB

Questions

Can Ornith 1.5 397B run on M5 Ultra 512 GB?

Ornith 1.5 397B at 6-bit runs on the M5 Ultra 512 GB with 256K of context at about 29 tokens a second.

Report VR-047QAY4, calculated Nov 22, 2026 from published specs, not benchmarked.