ax-engine 5.3.0

Python bindings for AX Engine

AX Engine

High-performance local inference engine for Apple Silicon — Python bindings.

Installation

Python (pip)

pip install ax-engine

Requires macOS 14+, Apple Silicon (M2 Max or newer), Python 3.10+.

Command-line tools (Homebrew)

To install the ax-engine-server HTTP adapter and ax-engine-bench CLI:

brew install defai-digital/ax-engine/ax-engine

Then verify:

ax-engine-server --help
ax-engine-bench doctor

Quick start

import ax_engine

session = ax_engine.Session(mlx=True, mlx_model_artifacts_dir="/path/to/model")
result = session.generate([token_id, ...], max_output_tokens=128)
print(result.output_tokens)

Or use the OpenAI-compatible shim:

python -m ax_engine.openai_server \
    --model-id my-model \
    --mlx-model-artifacts-dir /path/to/model \
    --tokenizer /path/to/tokenizer.json \
    --port 8080

Then point any OpenAI client at http://127.0.0.1:8080.

Optional dependencies

pip install "ax-engine[openai]"   # FastAPI + uvicorn for the OpenAI shim
pip install "ax-engine[download]" # mlx-lm for model downloading helpers

Requirements

• macOS 14 (Sonoma) or later
• Apple Silicon — M2 Max / M2 Ultra / M3 / M4 family (32 GB RAM minimum)
• Python 3.10+

Source

github.com/defai-digital/ax-engine