Use cases · Research organizer
Research often begins with scattered signals: an article, a social discussion and a recorded explanation. Savebase keeps that material together around the question you are trying to answer. Add evidence notes, label the source type and preserve the link behind each observation. When it is time to write a brief, you can return to the sources that support it instead of reconstructing the search.
Start with Free. Advanced feature availability depends on your plan.
Put your saves to work
Create a project folder such as Why users abandon setup, with tags for evidence, counterexample and open-question. Save first-hand sources alongside useful commentary, but label the distinction. Add the publication date and context to your notes when they matter. A clear question makes it easier to decide whether a new reference belongs in the collection.
Save selected web text through the extension's context menu with its source URL. Add a note explaining what the passage supports and what it leaves unresolved. On Pro, request transcripts for supported saved videos when spoken explanations are part of the evidence. Check the original recording for context before treating a transcript passage as a direct quotation.
With Pro, connect Savebase to ChatGPT or an MCP-compatible assistant. Ask it to find sources in the research folder, retrieve notes and available transcripts, and group observations with source links. Keep unverified suggestions separate from the evidence. MCP access is read-only, so changes to the research library still happen in Savebase when you choose to make them.
Start with a few references for something you want to do.
Savebase quick start →Write one research question and create a folder for its source material.
Save a small set of sources, attaching dates, relevant passages and evidence notes.
Compare the references manually or with a connected assistant, then check each source before drafting conclusions.
Optional AI workflow · Pro
Connect Savebase to ChatGPT or an MCP-compatible assistant, then adapt this prompt to your own folder and tags.
Find my saved sources in Why users abandon setup. Summarize the evidence and counterexamples with source URLs, use available notes and transcripts, and identify questions the saved material cannot answer.
Request video transcripts before asking about spoken content. MCP access is read-only; your assistant can retrieve available material, while library edits happen in Savebase.
Connect your AI assistant →Savebase organizes research references. Citation formatting, statistical analysis and source verification belong in your research process and specialist tools.
Begin with Free for web references, notes, tags and folders. Pro adds transcripts, ChatGPT/MCP access and integrations with Notion and Google Sheets.
See plans and current limits →Use a folder for each research question, tags for evidence types and notes for dates, definitions and caveats. Keep the original source URL attached to every useful observation. Review the collection before drafting so unsupported claims stay visible as open questions.
On Pro, a connected assistant can search and retrieve Savebase references, notes and available video transcripts through read-only MCP access. Ask for source-linked comparisons and gaps in the material. Check the original sources before relying on the resulting interpretation.
Build a source collection for your next newsletter, brief or article.
Explore the workflowSave selected passages with their URLs and add the context you want to remember.
Explore the workflowBuild a personal reference library you can search, review and use with AI.
Explore the workflowStart with one project and the references you want to return to.
Start your Savebase library