Skip to content

Mistral Small 4 119B on 4× RTX 3090: VRAM, speed and fit

Mistral Small 4 119B at 5-bit runs on the 4× RTX 3090 with 256K of context at about 109 tokens a second.

Text generation · Chat · Ollama / LM Studio

Runs comfortably · on 4× RTX 3090 · 5-bit (Q5_K_M) · Ollama / LM Studio · Chat

GPU memory
87.5 GB
of 96.0 GB
Context
256K
auto
Speed
109 tok/s
54 when full
First reply
0.4 s
2K prompt

Where the memory goes

Model
79.1 GB
Context
5.6 GB
Overhead
2.7 GB
OS
0.0 GB
Free
8.5 GB
Memory
87.5 GB of 96.0 GB
Context
256K tokens · auto
Speed
109 tok/s · 54 when full
First reply
0.4 s for a 2K prompt
Running cost
$20/mo at home · $131/mo on a cloud 4× RTX A5000

Validation: Validated

What Mistral Small 4 119B on 4× RTX 3090 was checked against
CheckTargetResultStatus
Fits in GPU memory96.0 GB usable87.5 GB neededPass
Headroom5% free8.5 GB free (9%)Pass
Context for chat16K or more256KPass
Writing speed15 tok/s or more109 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
Mistral Small 4 119B
Parameters
119B, 6.5B active
Precision
5-bit (Q5_K_M)
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
$131/mo
4× RTX A5000 at $1.08/h

4 h a day, power at $0.18 per kWh, cloud at RunPod on-demand prices as of Nov 21, 2026.

How to run it

llama.cpp
llama-server -hf unsloth/Mistral-Small-4-119B-2603-GGUF
vLLM
vllm serve mistralai/Mistral-Small-4-119B-2603
MLX (Mac)
mlx_lm.chat --model mlx-community/Mistral-Small-4-119B-2603-4bit

Mistral Small 4 119B on similar machines

All machines: Mistral Small 4 119B VRAM requirements. Everything that runs here: what can the 4× RTX 3090 run?

Other models that run on 4× RTX 3090

Questions

Can Mistral Small 4 119B run on 4× RTX 3090?

Mistral Small 4 119B at 5-bit runs on the 4× RTX 3090 with 256K of context at about 109 tokens a second.

Report VR-1UTG0XS, calculated Nov 22, 2026 from published specs, not benchmarked.