Inline AI

Inline AI covers the AI features that appear directly in the editor while you write code.

Autocomplete

AI autocomplete suggests code as you type. Enable it from Settings > AI > Autocomplete.

You can choose:

• OpenRouter: Athas loads the model list from the live Athas/OpenRouter model endpoint
• Custom: Athas sends autocomplete requests to an OpenAI-compatible endpoint you configure

When OpenRouter model loading fails, Athas leaves the model selector empty and lets you retry. It does not show hardcoded fallback models as selectable options.

Inline Edit

Inline edit uses the autocomplete model configuration to rewrite or generate code from the editor. Use it for focused edits where you want to keep the change close to the selected code.

Custom Autocomplete Endpoint

Select Custom in Settings > AI > Autocomplete Provider when you want completions or inline edit requests to go to your own endpoint.

You need:

• A base URL, such as http://localhost:11434/v1
• A model ID exactly as your endpoint expects it
• An optional API key if the endpoint requires a bearer token

The endpoint must expose an OpenAI-compatible chat completions API.