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.

SnapXO (Snapchat Export Organizer)

Turn a Snapchat data export into a clean, browsable archive with sorted media, efficient video encoding, GPS metadata, chat histories, statistics and an interactive map.

XO as in the prehistoric kiss emoji hihi^^ with X for eXport and O for Organizer.

pip install snapxo

Features

• Media Organization: Sort memories + chat media into structured folders with clean filenames (2026-05-08_0444.mp4)
• H.265 Encoding: Re-encode videos with Intel QSV hardware acceleration (auto-fallback to libx265)
• Overlay Burning: Burn Snapchat overlays onto photos/videos
• GPS/EXIF: Write GPS coordinates from memories_history.json into image EXIF data
• Duplicate Removal: MD5-based deduplication before encoding
• Voice Messages: Detect audio-only MP4s and convert to MP3
• Conversations: Per-contact chat pages with chat messages, photos, videos and voice messages
• Statistics: Overview page with cards + detail tables
• Snap Map: Interactive map with route, timeline slider, playback animation and Locations of each Snapchat Memories
• Media Gallery: index.html with navigation, type filters, thumbnails and a details panel per file (like Snapchat's Memories Tab)
• Merge: Combine several exports into one (deduplicated and renumbered, without re-encoding)
• PDF Export: Convert conversations and stats to PDF
• Resume: An interrupted run picks up where it stopped instead of copying and encoding everything again

Runs on Windows, Linux and macOS.

Installation

pip install snapxo

If something is missing, SnapXO says so and prints the install command for the system you are on.

For PDF export

playwright install chromium

This downloads the browser that renders the PDFs. On Linux it may also need system libraries, which playwright install-deps chromium takes care of.

For video processing

ffmpeg and ffprobe are needed for video encoding, overlay burning and voice conversion. They are native programs, so pip cannot install them:

If you cannot install one system-wide, pip install "snapxo[ffmpeg]" pulls in static-ffmpeg, which provides both binaries and is picked up automatically. Those builds normally don't ship with hardware encoders like Intel QSV, so encoding falls back to software and gets considerably slower. Choose a system-wide ffmpeg installation if possible.

Usage

# Basic: organize everything from ZIP(s)
snapxo mydata.zip -o ./output -y

# Multiple ZIPs (Snapchat splits large exports)
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

# Already extracted directory
snapxo ./extracted-export/ -o ./output -y

# Just show what's in the export
snapxo mydata.zip -o ./output --info

# Dry run (show what would happen)
snapxo mydata.zip -o ./output --dry-run

-o is required. The directory is created if it doesn't exist and reused if it does. --info and --dry-run still want the flag, but neither one creates the folder.

Also available as python -m snapxo ....

Passing a ZIP explicitly and passing a folder behave differently on purpose: a named ZIP that isn't a Snapchat export is an error, while ZIPs discovered by scanning a folder are silently skipped.

Custom ffmpeg path (e.g. WinGet install)

ffmpeg is located on PATH; you can also point at it explicitly with --ffmpeg-path and --ffprobe-path.

snapxo mydata.zip -o ./output -y \
  --ffmpeg-path "C:\...\ffmpeg.exe" \
  --ffprobe-path "C:\...\ffprobe.exe"

Re-generate specific outputs from an existing export

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

# Re-generate conversations as PDF
snapxo ./output/_meta/json -o ./output -y --only-conversations --conversation-format pdf

# Re-generate stats
snapxo ./output/_meta/json -o ./output -y --only-stats

# Re-generate map
snapxo ./output/_meta/json -o ./output -y --only-map

# Everything that can be a PDF, in one go: conversations + stats
snapxo ./output/_meta/json -o ./output -y \
  --only-conversations --only-stats --conversation-format pdf

Merging several output folders

# Merge two or more finished exports
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 conversations, stats, map and index are rebuilt from the combined JSON data. Media is copied, never re-encoded.

--hardlink links instead of copying: instant and free, but only works within one drive, and the merged files stay physically identical to the originals, so re-encoding the merged folder later would change the source folders too. --delete-sources removes the input folders, but only after re-hashing every merged file and asking for confirmation.

Output Structure

output/
├── 2022/                    folders with organized media
├── 2023/
├── ...
├── 2026/
│   ├── 2026-05-08_0444.mp4  H.265 encoded, with EXIF/GPS
│   ├── 2026-05-08_0445.jpg
│   ├── 2026-05-08_0446.mp3
│   └── ...
├── _stickers/               Custom sticker PNGs (usually empty)
├── _overlays/               Overlays that matched no media
├── _meta/                   Raw JSON + HTML from export
│   ├── manifest.json        What each file is and where it came from
│   ├── json/
│   └── html/
├── conversations/           Per-contact chat files
│   ├── john-doe.html
│   ├── john-doe.pdf
│   ├── max_mustermann.html
│   ├── group_my-group-chat.html
│   └── ...
├── index.html               Media gallery
├── stats.html               Statistics overview
├── stats.pdf
└── map.html                 Interactive Snap Map

Flags

Input and output

Mode

Filter (what to process)

Skip (what NOT to do)

Encoding

Output Format

Conversations

Stats

System

Merge

Resuming

Encoding a large export could take hours, so an interrupted run does not start over. While it works, SnapXO writes .snaporganizer_checkpoint.json into the output folder, recording which files it has already copied, converted, burned and encoded, and which steps are finished. Start the same command again and it picks up there:

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

Videos that are already encoded are not encoded again, overlays are not burned a second time (which would stack them on top of each other), and organized files are not overwritten with their unprocessed originals.

The checkpoint belongs to one specific run: it stores a fingerprint of the input, the output folder and every flag that changes what gets produced. Change any of them and the next run ignores it and starts fresh. --no-resume ignores it too.

Once a run completes, the checkpoint is deleted, so a finished export never carries one around. It also survives being killed mid-write, since it is written to a temporary file and moved into place.

Stickers

Snapchat exports sticker metadata in custom_sticker.json but does not include the PNG files, so the tool reports 0 of N sticker files found. Nothing can be done about that.

Privacy

SnapXO runs entirely on your machine. It reads the export ZIP you downloaded yourself, writes files to the folder you choose, and makes no network requests.

The browser used for PDF export renders local files only and is explicitly 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 in a browser, it loads its map library and map tiles from the internet. Generating it does not.

License

GPL-3.0