Query your GPT, Claude & Grok archives — find and read your old AI chats offline in the browser

What this app does

This is a private, browser-based tool for turning your exported AI chat history into a searchable personal archive. You import your conversation exports from ChatGPT (OpenAI), Claude (Anthropic), Grok (xAI), and Google Gemini, and the app builds a clean, readable, fully searchable library of every conversation — so you can find, re-read, and reuse things you said and things the AI said months ago. Everything runs locally in your browser: your chats are never uploaded to a server, there is no account, and there is no tracking.

The problem it solves

When you export your conversation history from AI chat tools like ChatGPT, Google Gemini, Anthropic Claude, or xAI Grok, the download is often a bundle of cryptic ".dat" files or dense ".json" export files instead of plain readable text. These files mix real conversation text together with binary noise — audio clips, images, and other attachments — and cram many separate conversations into one blob. There is no easy, private way to turn that messy export back into clean text you can actually read and search.

How it works

  1. You drop your ".dat" files, ".json" chat exports, or a ".zip" archive that contains them onto the page. Everything is read directly in your browser and nothing is uploaded to a server.
  2. The tool detects the text encoding (UTF-8 or UTF-16, with or without a byte-order mark) and decodes the readable text, while automatically detecting and skipping binary and audio files such as MP3, WAV, OGG, FLAC, MP4/M4A, PNG, JPEG, GIF, PDF, and ZIP.
  3. Each ".json" export (ChatGPT, Claude, Grok, Gemini) is split into one readable transcript per conversation, with the title, provider, date, and message-by-message "You:" / assistant labels preserved. Unrelated JSON in an export (billing, settings, memories) is skipped with a reason.
  4. Every converted conversation is saved to a persistent local library in your browser (IndexedDB), de-duplicated by content so importing the same export twice never creates duplicates. Your archive builds up over time across imports.
  5. You search your whole archive: full-text keyword search across every stored conversation, fuzzy find by title, provider filtering, and sorting by title, provider, model, date, message count, or word count.
  6. You open any conversation in a reader that renders it cleanly (including code blocks and tables) and lets you search within the transcript, with exact and fuzzy matching, match highlighting, and jump-to-next/previous navigation — search reaches inside code blocks too.
  7. You export what you find: download any conversation as a ".txt" file, download everything as a single ".zip", or export the metadata of your whole library as a ".csv".

Features

  • Imports ".dat" text files, ".json" chat exports (ChatGPT, Claude, Grok, Gemini), and ".zip" archives, via drag-and-drop or file browsing.
  • Splits multi-conversation exports into one clean transcript per conversation.
  • Persistent local library (IndexedDB) that grows across imports, de-duplicated by content.
  • Full-text search across your entire saved chat history, not just the latest import.
  • Fuzzy title find, provider filtering, and multi-field sorting.
  • In-transcript reader with exact and fuzzy in-page search, match highlighting, next/previous navigation, and search that includes code blocks.
  • Smart encoding detection for UTF-8 and UTF-16 text, including BOM handling.
  • Automatic skipping of binary and audio files, with a clear reason for each skip.
  • Card and table views, per-file download, download-all-as-zip, and CSV metadata export.
  • Light and dark mode.

Benefits

  • Completely private: everything runs locally in your browser, so your data never leaves your machine and there is no upload, account, or server storage.
  • Free and offline-capable once the page has loaded.
  • Turns unusable AI export bundles into a clean, permanent, searchable archive of your own AI conversations.
  • Lets you rediscover and reuse old prompts, answers, and ideas across ChatGPT, Claude, Grok, and Gemini from one place.
  • Fast, with no waiting on a server and no file-size limits beyond your own device.