Setup, troubleshooting, and reports
AI AdBlocker Support
Start with a narrow selector and one clear condition. Test it on the target page before expanding the rule.
Installation requires desktop Chrome 149 or later. Built-in AI availability can also vary by device and browser configuration.
First rule
Configure the model, scope, and match
- 01
Choose a provider
Use Chrome built-in AI, a local LM Studio model, or your own OpenAI or OpenRouter account.
- 02
Limit the candidates
Set a domain, optional path, and CSS selector before adding an AI condition.
- 03
Test the result
Use the rule test to see what matched, then adjust the instruction or threshold.
Quick diagnosis
Check these six things first
- 1
Confirm AI AdBlocker and the individual rule are enabled.
- 2
Check that the rule domain and optional path match the current page.
- 3
Use a CSS selector that matches the intended candidate elements.
- 4
Choose a compatible model for the embedding, prompt, or vision condition.
- 5
Add the required OpenAI or OpenRouter key when using a cloud model.
- 6
Run the built-in rule test, then reload the target page after changes.
If a local model fails, confirm LM Studio is running and its selected model is loaded. If a cloud model fails, check the key, provider balance or quota, and the selected model’s text or vision capability.
Reproducible reports
Share a minimal safe example
Never post API keys, private page text, personal information, or unredacted screenshots. A public test page and reduced rule are easier to diagnose.
- Environment
- Chrome version, extension version, operating system, provider, and model.
- Rule
- Domain, path, selector, condition type, and a sanitized instruction.
- Result
- Steps, expected result, actual result, and exact error text with secrets removed.
Still need help?
Use GitHub for public bugs and feature requests. Email the developer for privacy questions or sensitive contact.