Sider
Last updated August 26, 2026
Connect Sider to RouterOne through its custom OpenAI endpoint.
Sider is a browser extension that provides chat, translation, and page summaries in a sidebar. It accepts a custom OpenAI address and key.
1. Install
Search for "Sider" in your browser's extension store, install it, and enable it.
2. Create an RouterOne key
Create one on the keys page in the console. The key remains viewable there after creation, so you do not need to rotate it just to retrieve it again.
3. Configure
Open Sider's settings, go to General → AI access → Service provider, choose OpenAI, and fill in:
| Field | Value |
|---|---|
| API Key | Your RouterOne key |
| Base URL | https://gateway.routerone.ai/v1 |
Run the connection test and look for the success message.
4. Check it works
Open the sidebar on any page and send a message, or ask it to summarize the current page.
Choosing models
Pick from Sider's model list. If it looks incomplete, use "Update list" to fetch it again.
Names are canonical IDs and are not limited to OpenAI's models. Claude, Gemini, and other compatible models can also be called over the OpenAI protocol. For parameter differences in cross-protocol calls, see protocol conversion compatibility.
Troubleshooting
The connection test fails
Check the Base URL is https://gateway.routerone.ai/v1, with the /v1. Then check the key was copied in full (leading or trailing whitespace is the usual culprit), is still enabled in the console, and that the workspace has balance.
The model list is incomplete
Use "Update list" to fetch it again. Sider requests RouterOne's /v1/models, so the result is the full set of models available to that key.
The extension is blocked from making requests
Some browsers place extra restrictions on cross-origin requests from extensions. Check that Sider has the permissions it needs in your extension manager, and that no privacy extension is blocking requests to the RouterOne domain.
Summarizing a long page fails
The page text exceeds the model's context window. Switch to a model with a larger one; each model's page states it.