Choose the Right Local AI Model

The best model depends on the task, supported inputs, and the resources available on your computer.

Steps

1. Identify whether the task needs text reasoning, document analysis, vision, or image generation.
2. Confirm that the selected model supports the required input type.
3. Use a smaller model for routine work and a stronger model for complex synthesis when hardware allows.
4. For image understanding, verify both vision support and delivery of the actual image payload.
5. Run a short test after changing models before beginning important work.

What success looks like

The model accepts the required input, completes within practical time and memory limits, and produces an answer appropriate for the task.

If a model is unavailable

• Check model-service status.
• Confirm that the model is installed and active.
• Review hardware and context requirements.
• Select a supported fallback model and repeat a short test.

Hardware requirements · Back to Help Center

Last reviewed: July 21, 2026. Applies to the FridayLocalAI Alpha release.