> ## Documentation Index
> Fetch the complete documentation index at: https://docs.enconvo.ai/llms.txt
> Use this file to discover all available pages before exploring further.

# MLX Local Models

> Use on-device language models on Apple Silicon.

The **MLX (mlx)** AI Model provider runs supported models on Apple Silicon through Enconvo’s local model runtime.

## Use a local model

1. Open **Settings → AI Model → MLX (mlx)**.
2. Open the model picker and select a supported model.
3. Download the required resources when prompted, allowing enough disk space and memory for the selected model.
4. Wait for setup to complete, then test a short text request.
5. Review **Temperature** and **Reasoning Effort** when supported by the selected model.

Model files need an initial download; subsequent inference uses the local runtime. Model availability and hardware support depend on your build.

## Other local capabilities

Enconvo also includes MLX providers for speech recognition, text-to-speech, embeddings, reranking, and OCR. Each capability has its own model selection. See the [provider catalog](/providers/catalog).

If loading fails, inspect the error, available memory and disk space, and the model’s hardware requirements. Start with a smaller model before increasing context or concurrency.