og.cc.cd / guides / export-claude-chat-history

How to Export Your Claude Chat History (and Read the conversations.json File)

By og.cc.cd · Published · Updated · 4 min read

TL;DR: Export Claude chats from Settings → Privacy → Export data; you get an emailed ZIP whose conversations.json you can open in our Chat Export tool to read, search and save as Markdown, PDF or Word.

Updated · All sources ↓

Claude can email you a complete copy of your conversations. This guide walks through the official export, explains what is inside the file, and shows how to turn it into something you can actually read, search and archive.

Open the Chat Export viewer →All guides

Who can export

According to Anthropic’s help center, data exports are available to individual users on the Free, Pro and Max plans. On Team and Enterprise plans you can export your own chats only after an organization Owner or Primary Owner turns on “Allow members to export their own data”; Primary Owners instead export the whole organization from Organization settings → Data and privacy. Exports cannot be started from the Claude iOS or Android apps.

Step by step

  1. Open Claude on the web (claude.ai) or in Claude Desktop and sign in.
  2. Go to Settings → Privacy.
  3. Click Export data.
  4. Wait for the email sent to your account address. It contains a download link.
  5. Download while signed in to the same account. The link expires 24 hours after delivery; if it does, simply request a new export.

If you don’t see the button on a Team or Enterprise plan, the export setting is probably off — ask an owner. For Team and Enterprise member exports, Anthropic notes that the export includes your chats but not the contents of files you uploaded.

What is in the ZIP

The archive contains JSON files with your account data and your conversations. The one you want is conversations.json: a single JSON array with one object per chat. Anthropic does not publish a formal schema, so field names can change, but current exports look like this (shortened):

[
  {
    "uuid": "…",
    "name": "Plan a 3-day Kyoto trip",
    "created_at": "2026-10-12T09:12:03Z",
    "updated_at": "2026-10-12T09:20:41Z",
    "chat_messages": [
      { "uuid": "…", "sender": "human",     "created_at": "…",
        "text": "Plan a relaxed 3-day trip…", "content": [ … ],
        "attachments": [], "files": [] },
      { "uuid": "…", "sender": "assistant", "created_at": "…",
        "text": "", "content": [ { "type": "text", "text": "Here is a relaxed plan…" } ] }
    ]
  }
]
  • name is the chat title shown in the sidebar.
  • sender is human or assistant.
  • Message text may be in text, in the content array of typed blocks, or both — a robust reader checks both.
  • attachments and files list file names; treat them as references, not as copies of the files.

Turn the export into something readable

A raw JSON file with hundreds of chats is not pleasant to read. Our AI Chat Export viewer opens the ZIP or conversations.json directly in your browser — nothing is uploaded — and lets you:

  • browse chats in a sidebar and search titles;
  • save a single chat as Markdown, Word (.docx) or PDF (through the browser’s print dialog);
  • export every chat at once to a set of Markdown or Word files.

Step by step in the viewer: open /chat-export/, drop the ZIP onto the drop zone (or click Choose file), pick a chat from the list, then click Markdown, Word or PDF. To archive everything, use All → .md or All → .docx. You can disconnect from the internet after the page loads and it still works, which is a quick way to confirm nothing is uploaded.

Do it yourself with a script

If you prefer the command line, this Python snippet writes one Markdown file per conversation. It uses only the standard library:

import json, re, pathlib

chats = json.load(open("conversations.json", encoding="utf-8"))
out = pathlib.Path("claude-md"); out.mkdir(exist_ok=True)

def text_of(m):
    t = m.get("text") or ""
    if not t:
        t = "\n\n".join(b.get("text", "") for b in m.get("content", []) if b.get("type") == "text")
    return t.strip()

for c in chats:
    title = c.get("name") or "Untitled"
    lines = [f"# {title}", ""]
    for m in c.get("chat_messages", []):
        who = "You" if m.get("sender") == "human" else "Claude"
        lines += [f"**{who}:**", "", text_of(m), ""]
    slug = re.sub(r"[^\w-]+", "-", title)[:60].strip("-") or c["uuid"]
    (out / f"{slug}.md").write_text("\n".join(lines), encoding="utf-8")

Searching a large archive

After a year of daily use an export can hold thousands of chats. A few habits make it useful rather than just large:

  • Search titles first (name), then full text. Claude’s auto-generated titles are usually descriptive enough to find a topic.
  • Filter by created_at to find “that conversation from March” without reading everything.
  • Export only the chats worth keeping to Markdown and put them in your notes app or a git repository; plain text stays searchable for decades.

Troubleshooting

  • No email yet: Anthropic notes there may be a small delay while the export is generated, and the email goes to the address associated with your account — check spam and any alias you signed up with.
  • Link expired: links last 24 hours; repeat the export to get a new one.
  • No Export button: on Team or Enterprise the member-export setting is off, and Primary Owners use organization export instead.
  • On mobile: run the export from the web app or Claude Desktop; the iOS and Android apps cannot start one.

Privacy checklist

  • The export contains everything you have typed into Claude. Store it like any other sensitive document.
  • Do not upload it to random “converter” websites. Tools that run locally — a script, or a browser page that does not send data to a server — are safer.
  • An export cannot be imported into another personal Claude account; Anthropic does not support migrating data between personal accounts. Moving into a Team or Enterprise workspace has its own migration path.

Using ChatGPT or Gemini as well? See exporting ChatGPT conversations.json and saving Gemini chats.

Sources

Vendor features and prices were checked against these official pages on 2026-10-18. Products change; if something here no longer matches, the official page is right.

Related guides