Platforms

Supported extraction targets

Every entry below is a real parser in this repository — the class is imported from the same subpath you use in your own project.

AI chat platforms

Platform Parser Strategy Notes
ChatGPT ChatGPTParser
decant-core/ai/chatgpt
DOM + internal API Fallback and RPC-assisted extraction; scroll/dedup helper for long threads.
Claude ClaudeParser
decant-core/ai/claude
DOM + React fiber Reads the React tree for artifacts and structured blocks with DOM fallback.
Google Gemini GeminiParser
decant-core/ai/gemini
DOM + internal API Uses batchexecute RPC pagination with resilient DOM fallback.
Microsoft Copilot CopilotParser
decant-core/ai/copilot
DOM Multi-domain (bing + copilot) DOM extraction.
Perplexity PerplexityParser
decant-core/ai/perplexity
Internal API + DOM Internal API first, DOM fallback for source citations and answers.
DeepSeek DeepSeekParser
decant-core/ai/deepseek
DOM + internal API API-assisted parsing with DOM fallback.
Qwen QwenParser
decant-core/ai/qwen
DOM Structured DOM extraction incl. file attachments.
Meta AI MetaParser
decant-core/ai/meta
DOM DOM extraction.
Mistral / Le Chat MistralParser
decant-core/ai/mistral
DOM DOM extraction.
Proton Lumo LumoParser
decant-core/ai/lumo
DOM DOM extraction.
Z.ai ZAiParser
decant-core/ai/z_ai
DOM DOM extraction.
Google AI Studio GoogleAIStudioParser
decant-core/ai/google_ai_studio
DOM DOM extraction for aistudio.google.com sessions.
NotebookLM NotebookLMParser
decant-core/ai/notebooklm
DOM DOM extraction incl. notes and citations.
Google Search AI (AI Overviews) GoogleSearchAIParser
decant-core/ai/google_search_ai
DOM DOM extraction for SGE overviews.
Gemini Cloud Assist GeminiCloudAssistParser
decant-core/ai/gemini_cloud_assist
DOM DOM extraction for console.cloud.google.com assistants.
Joyland JoylandParser
decant-core/ai/joyland
DOM DOM extraction for character chat platforms.
Chub ChubParser
decant-core/ai/chub
DOM DOM extraction.

Web articles

Target Parser Strategy Notes
Any web article ArticleParser
decant-core/article
Readability + Defuddle + Article-Extractor Runs three extractors concurrently and arbitrates by content-quality scoring.
Maintenance is the point. “DOM” parsers target semantic structure, not pixel positions; “internal API” parsers read the same RPC payloads the frontend uses. When a platform changes, one shared fix heals every consumer. If a parser drifts, add a fixture and open a PR — see Contributing.