Microsoft · First MLX build 27 Feb 2025 · In Core AI since 20 Aug 2026
Phi-4-mini-instruct
MLXCore AI
Text mit Trending
A 3.8B parameter text generation model from Microsoft, with 7 MLX builds, from 4-bit to fp16 and Apple Core AI export recipes for macOS.
- Parameters
- 3.8B From Hugging Face
- Quantisation
- 4-bit Phi-4-mini-instruct-4bit
- Memory needed
- 5 GB Estimate, unified memory
- Download
- 2.2 GB Weights, estimated
- Downloads
- 2.7K This build, last 30 days
- Task
- Text generation Text
- MLX builds
- 7 On mlx-community
- Devices
- Mac Apple silicon
- Parameters
- 3.8B From Hugging Face
- Compression
- 4-bit Apple’s preset
- Memory needed
- 5 GB Estimate from parameters
- Context
- 128K tokens Fixed at export
- Precision
- FP16 Compute precision
- Platforms
- macOS Mac
- Export presets
- 1 In Apple’s registry
- Listed by Apple
- 20 Aug 2026 First registry commit
Memory
Will it fit?Needs about 5 GB at 4-bit
8 GB
16 GB
24 GB
32 GB
64 GB
128 GB
Fits with headroomTight, close other appsWon’t fit
Estimated from the size of the weights, plus a fifth for context and working memory and 2 GB for macOS and your apps. Long contexts need more.
Run it
Terminal · mlx-lm$ pip install -U mlx-lm
$ mlx_lm.generate --model mlx-community/Phi-4-mini-instruct-4bit --prompt "Hello"Memory
Will it fit?Needs about 5 GB at 4-bit
8 GB
16 GB
24 GB
32 GB
64 GB
128 GB
Fits with headroomTight, close other appsWon’t fit
Estimated from the parameter count at Apple’s compression setting. Apple does not publish package sizes.
Export it
Terminal · uv$ git clone https://github.com/apple/coreai-models
$ cd coreai-models
$ uv run coreai.llm.export phi-4-mini-instruct
Apple publishes the recipe, not a ready-made package: these commands convert the original weights to a Core AI .aimodel on your Mac. See the repository for requirements.
MLX builds
7 on mlx-community| Phi-4-mini-instruct-4bit | 4-bit | 2.2 GB | 5 GB | 2.7K |
|---|---|---|---|---|
| Phi-4-mini-instruct-mlx-4Bit | 4-bit | 2.2 GB | 5 GB | 480 |
| Phi-4-mini-instruct-6bit | 6-bit | 3.1 GB | 6 GB | 241 |
| Phi-4-mini-instruct-mlx-6Bit | 6-bit | 3.1 GB | 6 GB | 135 |
| Phi-4-mini-instruct-8bit | 8-bit | 4.1 GB | 7 GB | 768 |
| Phi-4-mini-instruct-mlx-8Bit | 8-bit | 4.1 GB | 7 GB | 252 |
| Phi-4-mini-instruct-mlx-fp16 | fp16 | 7.7 GB | 11 GB | 224 |
Sizes are estimated from Hugging Face’s parameter counts and the quantisation. Downloads are for the last 30 days.
Core AI export presets
From Apple’s registry| Preset | Platform | Compression | Precision | Context |
|---|---|---|---|---|
| phi-4-mini-instruct | macOS | 4-bit | FP16 | 128K |