SnapXO (Snapchat Export Organizer)

Snapchat hands you your data as a few thousand files named by ID and a pile of JSON. This turns it into an archive you can still open in ten years: a clean, browsable archive with sorted media, efficient video encoding, GPS metadata, chat histories, statistics and an interactive map.

pip install snapxo

Features

• Media Gallery: index.html with navigation, type filters, thumbnails and a details panel per file
• Chat Overview: chats.html lists every chat like the app does, with a search across all messages
• Conversations: per-contact chat pages with messages, photos, videos and voice messages
• Media Organization: memories and chat media in structured folders with clean filenames (2026-05-08_0444.mp4)
• H.265 Encoding: Intel QSV hardware acceleration, automatic fallback to libx265
• Overlay Burning: Snapchat overlays burned onto photos and videos
• GPS/EXIF: coordinates from memories_history.json written into image EXIF
• Duplicate Removal: MD5 deduplication, before encoding
• Voice Messages: audio-only MP4s detected and converted to MP3
• Statistics: memories, chat media, friends, calls, searches and your Snapscore, with a table behind every number
• Snap Map: every memory pinned where it was taken, with a timeline you can play back
• Merge: several exports combined into one, deduplicated and renumbered, without re-encoding
• PDF Export: conversations, statistics and the media gallery as PDF
• Verify: a finished archive checked against its manifest and its checksums
• Resume: an interrupted run picks up where it stopped

Runs on Windows, Linux and macOS.

Installation

pip install snapxo

snapxo doctor says what is missing and prints the install command for the system you are on.

For PDF export: playwright install chromium downloads the browser that renders them. On Linux it may also need playwright install-deps chromium.

For video processing: ffmpeg and ffprobe are native programs, so pip cannot install them.

pip install "snapxo[ffmpeg]" provides both binaries through static-ffmpeg if you cannot install them system-wide. Those builds have no hardware encoding, so prefer a system-wide ffmpeg.

Usage

# Organize everything from ZIP(s)
snapxo mydata.zip -o ./output -y
snapxo export1.zip export2.zip -o ./output -y

# A folder full of ZIPs, anything that isn't a Snapchat export is ignored
snapxo ~/Downloads -o ./output -y

# An already extracted directory
snapxo ./extracted-export/ -o ./output -y

# Look before you write
snapxo mydata.zip --info
snapxo mydata.zip --dry-run

# One year only, and the gallery as a PDF as well
snapxo mydata.zip -o ./output -y --since 2026-01-01 --until 2026-12-31
snapxo mydata.zip -o ./output -y --index-format pdf

-o is required whenever something is written. The directory is created if it doesn't exist and reused if it does. --info and --dry-run write nothing, so they work without it. Everything is also available as python -m snapxo ....

A named ZIP that isn't a Snapchat export is an error, while ZIPs found by scanning a folder are skipped silently. ffmpeg is taken from PATH unless --ffmpeg-path and --ffprobe-path point elsewhere.

Re-generate pages from a finished export

The input is the _meta/json folder of the finished export, not the export.

snapxo ./output/_meta/json -o ./output -y --only-conversations --conversation-format pdf
snapxo ./output/_meta/json -o ./output -y --only-stats
snapxo ./output/_meta/json -o ./output -y --only-map

Check a finished archive

# Seconds: is every file from the manifest there, and the right size
snapxo verify ./output

# Reads everything and compares it to _meta/checksums.json, the first run writes it
snapxo verify ./output --hash

Worth doing after moving the archive to another drive, and once in a while against bit rot. --update accepts the current state as the new baseline, and --checksums writes the baseline during a normal run.

Merge several output folders

snapxo merge ./out-2024 ./out-2026 -o ./merged

# Or point at a folder that contains them all
snapxo merge ./all-exports -o ./merged

# Instant merge without extra disk space, then drop the originals
snapxo merge ./all-exports -o ./merged --hardlink --delete-sources

Duplicates are removed by content hash, everything is renumbered chronologically, and the pages are rebuilt from the combined JSON data. Media is copied, never re-encoded. --hardlink is instant and free but works within one drive only, and the merged files stay physically identical to the originals. --delete-sources removes the inputs, after re-hashing every merged file and asking.

Every input folder is checked against its manifest first, and against its checksums where they exist. A damaged file stops the merge. --no-verify takes those files along instead, marked in the manifest, in _meta/integrity.json, on the gallery tile and at the attachment in the conversation, and --skip-damaged leaves them out. If another export holds the same file intact, the intact copy wins. --delete-sources cannot be combined with --no-verify.

Output Structure

output/
├── 2022/                    folders with organized media
├── ...
├── 2026/
│   ├── 2026-05-08_0444.mp4  H.265 encoded, with EXIF/GPS
│   ├── 2026-05-08_0445.jpg
│   └── 2026-05-08_0446.mp3
├── _overlays/               Overlays that matched no media
├── _meta/                   Raw JSON + HTML from export
│   ├── manifest.json        What each file is and where it came from
│   ├── checksums.json       Fingerprint for `snapxo verify`
│   ├── integrity.json       Files that arrived damaged, if any
│   ├── thumbs/              Preview images for the gallery and the chats
│   ├── json/
│   └── html/
├── conversations/           Per-contact chat files
│   ├── john-doe.html
│   ├── john-doe.pdf
│   └── group_my-group-chat.html
├── index.html               Media gallery
├── index.pdf
├── chats.html               Chat overview with search
├── stats.html               Statistics overview
├── stats.pdf
└── map.html                 Interactive Snap Map

Flags

Input and output

Filter (what to process)

Skip (what NOT to do)

Encoding

Output Format

Conversations and stats

System

Merge

Verify

Resuming

Encoding a large export can take hours, so an interrupted run does not start over. SnapXO writes .snaporganizer_checkpoint.json into the output folder while it works and picks up there when the same command runs again:

Resuming from checkpoint (1 steps, 14 files already done)

Encoded videos are not encoded twice, overlays are not burned on top of each other, and organized files are not overwritten with their originals. The checkpoint stores a fingerprint of the input, the output folder and every flag that changes what gets produced, so changing any of them starts fresh, as does --no-resume. A completed run deletes it.

Stickers

Snapchat exports sticker metadata in custom_sticker.json but none of the PNG files, and STICKER messages carry no sticker reference. There is nothing to export, which is why 1.1.0 dropped the step.

Privacy

SnapXO runs entirely on your machine, on the export you downloaded yourself, and makes no network requests. The browser used for PDF export is put into offline mode, so it cannot reach the network even if a page asked it to.

The one exception is map.html: when you open it, it loads its map library and map tiles from the internet. Generating it does not.

Not affiliated with Snap Inc. This is an independent, unofficial open-source tool. It is not created, endorsed, sponsored or supported by Snap Inc., and has no connection to Snapchat, Snap Inc. or any of its products. "Snapchat" and "Snap" are trademarks of Snap Inc., used here only to describe what this tool reads. SnapXO works purely offline on a data export you requested and downloaded yourself. This tool never logs into an account and never contacts Snapchat.

License

GPL-3.0