Skip to content

Ornith 1.5 35B A3B on 2× RTX 5090: VRAM, speed and fit

Ornith 1.5 35B A3B at 8-bit runs on the 2× RTX 5090 with 256K of context at about 200 tokens a second.

Text generation · Chat · Ollama / LM Studio

Runs comfortably · on 2× RTX 5090 · 8-bit (Q8_0) · Ollama / LM Studio · Chat

GPU memory
42.3 GB
of 64.0 GB
Context
256K
auto
Speed
200 tok/s
108 when full
First reply
0.1 s
2K prompt

Where the memory goes

Model
35.6 GB
Context
5.1 GB
Overhead
1.7 GB
OS
0.0 GB
Free
21.7 GB
Memory
42.3 GB of 64.0 GB
Context
256K tokens · auto
Speed
200 tok/s · 108 when full
First reply
0.1 s for a 2K prompt
Running cost
$17/mo at home · $66/mo on a cloud 2× RTX A5000

Validation: Validated

What Ornith 1.5 35B A3B on 2× RTX 5090 was checked against
CheckTargetResultStatus
Fits in GPU memory64.0 GB usable42.3 GB neededPass
Headroom5% free21.7 GB free (34%)Pass
Context for chat16K or more256KPass
Writing speed15 tok/s or more200 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 1.5 35B A3B
Parameters
36B, 3B active
Precision
8-bit (Q8_0)
Context
256K, auto
Chats at once
1
Use
Chat
Runs in
Ollama / LM Studio
Hardware
2× RTX 5090
GPU can use
64.0 GB of 64.0 GB

Running cost

At home
$17/mo
770 W at the wall
Renting instead
$66/mo
2× RTX A5000 at $0.54/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 peculiar-ragdoll/Tiel-Coder-35B-A3B-GGUF-MTP
vLLM
vllm serve ornith-ai/Ornith-1.5-35B-A3B
MLX (Mac)
mlx_lm.chat --model mlx-community/Ornith-1.5-35B-A3B-OptiQ-4bit

Ornith 1.5 35B A3B on similar machines

All machines: Ornith 1.5 35B A3B VRAM requirements. Everything that runs here: what can the 2× RTX 5090 run?

Other models that run on 2× RTX 5090

Questions

Can Ornith 1.5 35B A3B run on 2× RTX 5090?

Ornith 1.5 35B A3B at 8-bit runs on the 2× RTX 5090 with 256K of context at about 200 tokens a second.

Report VR-04WI8K4, calculated Nov 22, 2026 from published specs, not benchmarked.