Skip to content

Ministral 3 14B on M3 Ultra 96 GB: VRAM, speed and fit

Ministral 3 14B at 8-bit runs on the M3 Ultra 96 GB with 256K of context at about 34 tokens a second.

Text generation · Chat · Ollama / LM Studio

Runs comfortably · on M3 Ultra 96 GB · 8-bit (Q8_0) · Ollama / LM Studio · Chat

GPU memory
78.5 GB
of 96.0 GB
Context
256K
auto
Speed
34 tok/s
9.1 when full
First reply
2.5 s
2K prompt

Where the memory goes

Model
13.8 GB
Context
40.0 GB
Overhead
672 MB
OS
24.0 GB
Free
17.5 GB
Memory
78.5 GB of 96.0 GB
Context
256K tokens · auto
Speed
34 tok/s · 9.1 when full
First reply
2.5 s for a 2K prompt
Running cost
$3.55/mo at home · $117/mo on a cloud 4× RTX 2000 Ada

Validation: Validated

What Ministral 3 14B on M3 Ultra 96 GB was checked against
CheckTargetResultStatus
Fits in GPU memory72.0 GB usable54.5 GB neededPass
Headroom5% free17.5 GB free (24%)Pass
Context for chat16K or more256KPass
Writing speed15 tok/s or more34 tok/sPass
First replyunder 3 s, 2K prompt2.5 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
Ministral 3 14B
Parameters
13.9B
Precision
8-bit (Q8_0)
Context
256K, auto
Chats at once
1
Use
Chat
Runs in
Ollama / LM Studio
Hardware
M3 Ultra 96 GB
GPU can use
72.0 GB of 96.0 GB

Running cost

At home
$3.55/mo
162 W at the wall
Renting instead
$117/mo
4× RTX 2000 Ada at $0.96/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

Ollama
ollama run ministral-3:14b
llama.cpp
llama-server -hf unsloth/Ministral-3-14B-Instruct-2512-GGUF
vLLM
vllm serve mistralai/Ministral-3-14B-Instruct-2512
MLX (Mac)
mlx_lm.chat --model mlx-community/Ministral-3-14B-Instruct-2512-4bit

Ministral 3 14B on similar machines

All machines: Ministral 3 14B VRAM requirements. Everything that runs here: what can the M3 Ultra 96 GB run?

Other models that run on M3 Ultra 96 GB

Questions

Can Ministral 3 14B run on M3 Ultra 96 GB?

Ministral 3 14B at 8-bit runs on the M3 Ultra 96 GB with 256K of context at about 34 tokens a second.

Report VR-0JUDO49, calculated Nov 22, 2026 from published specs, not benchmarked.