Skip to content

Qwen3.8 27B on 4× RTX 3090: VRAM, speed and fit

Qwen3.8 27B at 8-bit runs on the 4× RTX 3090 with 256K of context at about 22 tokens a second.

Text generation · Chat · Ollama / LM Studio

Runs comfortably · on 4× RTX 3090 · 8-bit (Q8_0) · Ollama / LM Studio · Chat

GPU memory
47.0 GB
of 96.0 GB
Context
256K
auto
Speed
22 tok/s
14 when full
First reply
0.5 s
2K prompt

Where the memory goes

Model
27.5 GB
Context
16.1 GB
Overhead
3.4 GB
OS
0.0 GB
Free
49.0 GB
Memory
47.0 GB of 96.0 GB
Context
256K tokens · auto
Speed
22 tok/s · 14 when full
First reply
0.5 s for a 2K prompt
Running cost
$20/mo at home · $66/mo on a cloud 2× RTX A5000

Validation: Validated

What Qwen3.8 27B on 4× RTX 3090 was checked against
CheckTargetResultStatus
Fits in GPU memory96.0 GB usable47.0 GB neededPass
Headroom5% free49.0 GB free (51%)Pass
Context for chat16K or more256KPass
Writing speed15 tok/s or more22 tok/sPass
First replyunder 3 s, 2K prompt0.5 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
Qwen3.8 27B
Parameters
27.8B
Precision
8-bit (Q8_0)
Context
256K, auto
Chats at once
1
Use
Chat
Runs in
Ollama / LM Studio
Hardware
4× RTX 3090
GPU can use
96.0 GB of 96.0 GB

Running cost

At home
$20/mo
920 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

Ollama
ollama run qwen3.8:27b
llama.cpp
llama-server -hf unsloth/Qwen3.8-27B-GGUF
vLLM
vllm serve Qwen/Qwen3.8-27B
MLX (Mac)
mlx_lm.chat --model mlx-community/Qwen3.8-27B-4bit

Qwen3.8 27B on similar machines

All machines: Qwen3.8 27B VRAM requirements. Everything that runs here: what can the 4× RTX 3090 run?

Other models that run on 4× RTX 3090

Questions

Can Qwen3.8 27B run on 4× RTX 3090?

Qwen3.8 27B at 8-bit runs on the 4× RTX 3090 with 256K of context at about 22 tokens a second.

Report VR-0Z8IAIW, calculated Nov 22, 2026 from published specs, not benchmarked.