> ## Documentation Index
> Fetch the complete documentation index at: https://ade-ac1c6011-ade-mobile-chat-stability-54bd7ef0.mintlify.site/llms.txt
> Use this file to discover all available pages before exploring further.

# Connect an AI provider

> Add one working provider so ADE can run coding agents.

Open **Settings -> AI Providers**. ADE needs at least one verified provider before it can run agent chat, workers, CTO delegation, auto-titles, or AI summaries.

## Common choices

| Provider                         | Best for                                              | Setup                                               |
| -------------------------------- | ----------------------------------------------------- | --------------------------------------------------- |
| **Claude / Claude Code**         | Strong default for coding tasks and project reasoning | API key or authenticated `claude` CLI               |
| **Codex**                        | OpenAI coding agent workflows                         | Authenticated Codex CLI or configured OpenAI access |
| **Cursor**                       | Cursor agent runs from inside ADE                     | Cursor API key                                      |
| **OpenCode**                     | Local or OpenRouter-backed CLI workflows              | OpenCode config and provider key                    |
| **OpenRouter / local providers** | Cost control or local experimentation                 | API key or local endpoint                           |

## Add and verify

<Steps>
  <Step title="Choose the provider">
    Pick the provider you already use, or start with Claude if you are unsure.
  </Step>

  <Step title="Add credentials">
    Paste the API key or make sure the provider CLI is installed and authenticated.
  </Step>

  <Step title="Click verify or refresh">
    ADE checks the provider and shows whether it can start sessions.
  </Step>

  <Step title="Set a default model">
    Choose a default model for chat. You can override it per session later.
  </Step>
</Steps>

<Warning>
  Provider calls go from your machine to the provider. ADE does not proxy model requests through an ADE-hosted service.
</Warning>

## If verification fails

* Confirm the key is active and has model access.
* Restart ADE after installing a new CLI.
* Open a terminal and test the provider CLI directly.
* Check **Settings -> AI Providers** again and click **Refresh**.