Next-level performance
Faster chip, more RAM, more storage — our most powerful Mac Mini configuration available.
Apple Mac mini M4 Pro
The power of the most advanced Mac mini, as dedicated servers designed for your demanding projects, including AI and machine learning.
Ready in a click
Launch your Mac mini M4 Pro in a click from our Public Cloud Console and access it right away through VNC or CLI.
Hosted in Europe
Rely on our Mac mini M4 Pro as a Service hosted in our European datacenter in Paris, France for a unique sovereign offer.
Technical specifications
- processor
CPU
Apple M4 Pro chip. 14-core CPU. 20-core GPU. 16-core Neural Engine.
- memory
Memory
64GB LPDDR5X
- storage
Storage
2TB
- storage_type
Disk type
SSD
- bandwidth
Bandwidth
Up to 10 Gbps
- service_level
OS
Latest macOS versions available
Mac mini M4 Pro for Artificial Intelligence
High-performance AI at lower cost
The M4-XL (64 GB) enables you to run models up to 24B (16 bits) for €0.49/h, whereas an L4 or L40S would cost two to three times more (up to €1.72/h).
Easy access
Mac mini M4 Pro is ideal for batch/CLI. With llm-mlx, directly download and launch models from the Hugging Face Hub using their ID.
Discover the performance
| Model | Input tokens/sec | Output tokens/sec | Peak memory (Gb) |
|---|---|---|---|
| 70B, 4 bitsmlx-community/Llama-3.3-70B-Instruct-4bit | 29.1 | 6.2 | 39.8 |
| 32B, 8 bitsmlx-community/DeepSeek-R1-Distill-Llama-8B-8bit | 75.7 | 30 | 8.7 |
| 8B, 4 bitsmlx-community/DeepSeek-R1-Distill-Llama-8B-4bit | 124.3 | 53.6 | 4.7 |
| 12B, 8 bitsmlx-community/Mistral-Nemo-Instruct-2407-8bit | 60.2 | 19.5 | 13.1 |
| 27B, 8 bitsmlx-community/gemma-2-27b-it-8bit | 15.6 | 8.4 | 29.5 |
| 32B, 8 bitsmlx-community/Qwen2.5-32B-Instruct-8bit | 48.2 | 7.1 | 34.9 |
Apple Intelligence at your fingertips
Apple frameworks
Train your models with MLX on Apple Silicon CPU & GPU and leverage the Foundation Models Framework to directly integrate Apple AI into your applications.
Daily boost
Execute LLM queries, summarize articles, structure your notes, or generate code in Xcode.
Integrated productivity
Intelligent assistants, autonomous agents, coding intelligence: Apple AI easily integrates into your tools.
A dedicated Apple server for heavy workloads
High performance
The M4 Pro chip provides significant improvements that makes it well-suited for demanding tasks such as machine learning, data processing, and 3D rendering.
Unified memory architecture
Apple’s unified memory system allows the CPU, GPU, and Neural Engine to access the same memory pool, which boosts performance and efficiency for AI workloads that require fast data exchange.
Optimized macOS ecosystem
macOS is increasingly optimized for AI and ML workflows, with native support for frameworks like Core ML, Create ML, and compatibility with popular libraries, or native Apple Silicon builds.
Full access and autonomy
All the performance dedicated to your use, with enhanced security fully in your hands. Customize settings, leverage tools, and manage your mac mini to suit your needs perfectly, optimizing its use to your preferences.
Want to discuss a big project with our Apple silicon?
Get started with tutorials
Configuring a GitHub Actions RunnerAccess the documentation- How to set up a VM with UTMAccess the documentation