# Continue (/docs/development/developer-tools/continue)



<Callout type="info" title="Continue is currently not supported">
  Continue does implement the OpenAI **Responses API**, but it only uses it for models whose names match OpenAI's own naming — `o1`/`o3`/`o4`-style or `gpt-5` and above. For every other model name it calls the legacy Chat Completions endpoint (`/v1/chat/completions`), which the CoreAI API does not implement, so requests fail with `404 Not Found` regardless of authentication settings. There is no configuration option to force the Responses API for other model names (`useResponsesApi` can only disable it, and the upstream feature request for OpenAI-compatible providers was closed as not planned). Since platform model names generally do not follow OpenAI's naming, Continue cannot reliably connect to the CoreAI API. Use [OpenCode](/docs/development/developer-tools/opencode) or the OpenAI SDK instead. This page is kept for reference and will be updated if Continue adds Responses API support for arbitrary models.
</Callout>

`Continue` would connect to BullSequana AI through the `CoreAI API`; the configuration below documents the intended setup, but it does not work until the compatibility gap described above is closed.

What To Use [#what-to-use]

Use:

* provider: `openai`
* base URL: `https://llm-backend.<platform-domain>/v1`
* token: JWT or `sk-bsq-...` API key

Do not point Continue directly to internal LiteLLM endpoints.

1\. Install Continue [#1-install-continue]

Install the Continue extension in:

* `VS Code`
* `JetBrains`

Example for VS Code:

```bash
code --install-extension Continue.continue
```

2\. Prepare Authentication [#2-prepare-authentication]

Generate or obtain one of these:

* a JWT token
* a BullSequana AI API key

If you need that flow first, use:

* [Use local models via API](/docs/development/use-local-models-via-api)

3\. Configure Continue [#3-configure-continue]

Create or edit `~/.continue/config.json`.

Example configuration:

```json
{
  "models": [
    {
      "title": "BullSequana AI",
      "provider": "openai",
      "model": "<model-name-from-/v1/models>",
      "apiKey": "sk-bsq-v1-...",
      "apiBase": "https://llm-backend.<platform-domain>/v1"
    }
  ],
  "tabAutocompleteModel": {
    "title": "BullSequana AI Fast",
    "provider": "openai",
    "model": "<fast-model-name-from-/v1/models>",
    "apiKey": "sk-bsq-v1-...",
    "apiBase": "https://llm-backend.<platform-domain>/v1"
  }
}
```

4\. Check Model Names [#4-check-model-names]

Do not hardcode example model names from another environment without checking first.

Available models are also visible in the `CoreAI Portal`.

If you want the API-level source of truth for the current environment, use:

```bash
curl "https://llm-backend.<platform-domain>/v1/models" \
  -H "Authorization: Bearer <your-token>"
```

5\. Start Using Continue [#5-start-using-continue]

Common shortcuts include:

* `Ctrl+L` for chat
* `Ctrl+I` for inline edits
* editor-native completion flows for code suggestions

Why This Is The Recommended Path [#why-this-is-the-recommended-path]

This keeps Continue aligned with the BullSequana AI platform contract:

* stable API surface
* controlled authentication
* flexibility to evolve internal routing and inference components later

Related Pages [#related-pages]

* [Developer Tools](/docs/development/developer-tools)
* [Use local models via API](/docs/development/use-local-models-via-api)
* [API Reference](/docs/api-reference)