VideoAdGuard is a browser extension designed to automatically detect and allow users to skip embedded advertisements within Bilibili (B站) videos. It leverages large language models (LLMs) to analyze video subtitles and identify ad content, offering a convenient way for users to bypass these interruptions.

How It Works

The extension operates by first obtaining the video's subtitle content, along with its title and pinned comments. This data is then sent to a configured LLM for analysis. The LLM's output is used to determine specific time segments within the video that are likely to contain advertisements. Upon detection, a "Skip Ad" button is displayed on the video player.

Quick Start & Requirements

• Installation:
  • Edge: Install directly from the Microsoft Edge Add-ons store.
  • Other Browsers (Tampermonkey): Install Tampermonkey, then install the script from GreasyFork.
  • Chrome-based Browsers (Manual): Download the release package, unzip, enable developer mode in extensions, and load the unpacked extension.
• Prerequisites: Requires a configured LLM API key and model. The extension functions best with videos that have subtitles.
• Configuration: Users need to set up their API key and model choice, supporting various LLMs including Zhipu AI, DeepSeek, OpenAI, and local Ollama instances.
• Documentation: Edge插件商店 - VideoAdGuard

Highlighted Details

• Utilizes LLMs to analyze subtitles for precise ad detection.
• Supports a wide range of LLM providers and local Ollama.
• Offers a one-click skip button for detected ad segments.
• Available as a direct browser extension or a Tampermonkey script.

Maintenance & Community

• Repository: Warma10032/VideoAdGuard
• Feedback: GitHub Issues are the primary channel for bug reports and suggestions.

Licensing & Compatibility

• License: GPL-2.0 License. This is a copyleft license, meaning derivative works must also be open-sourced under the same license.
• Commercial Use: The GPLv2 license may impose restrictions on linking with closed-source commercial applications.

Limitations & Caveats

The plugin requires video subtitles to be present for functionality. The effectiveness of ad detection is directly dependent on the capability of the chosen LLM. Users should be mindful of token consumption when using paid API services.