You can keyword 100 stock photos in under an hour. The process: open ShotMeta, point it at your folder, and let it analyse each image with AI to generate a title, description, and keyword set that meets Adobe Stock, Shutterstock, and Getty submission requirements. Your files never leave your computer. No cloud upload, no subscription, no privacy risk. This guide walks through the exact workflow, step by step.
In this guide
- Why keywording stock photos takes so long
- What submission-ready metadata actually needs
- How ShotMeta analyses your images without uploading them
- Step-by-step: keyword 100 photos with ShotMeta
- How long does it actually take?
- ShotMeta vs other AI metadata tools
- FAQ
Why does keywording stock photos take so long?
Manual keywording is the bottleneck most stock photographers never fully solve. A single image needs a descriptive title (under 200 characters), a natural-language description, and 30 to 50 relevant keywords ranked from most to least specific. Multiply that by a shoot of 100 images and you are looking at 4 to 8 hours of repetitive work before you have uploaded a single file.
The alternatives most photographers try first have their own problems. Generic keyword lists miss the specific subjects and moods that drive search traffic on microstock platforms. Browser-based AI tools require you to upload your photos to a third-party server, which creates a real privacy problem if you shoot identifiable subjects, locations, or unreleased products. And most tools do not write descriptions in the format each platform rewards.
AI-powered metadata for stock photos and videos. Runs locally on your machine, no cloud upload required. Works with Adobe Stock, Shutterstock, Getty, and Pond5.
What does submission-ready stock photo metadata actually need?
Every major microstock platform looks for the same three components, in slightly different formats:
- Title: 5 to 10 words, sentence case, descriptive of the main subject. No keyword stuffing.
- Description: 1 to 3 sentences describing the subject, mood, setting, and potential use case. Platforms use this for search ranking and buyer preview text.
- Keywords: 30 to 50 terms, comma-separated, ordered from most to least relevant. Include the subject, action, emotion, colour, location type, and use case. Avoid repeating words already in the title.
The goal is metadata a buyer would actually search for, not a list of every object in the frame. A photo of two people shaking hands in an office needs keywords like "business agreement", "professional handshake", and "corporate partnership" before it needs "hand" and "office".
How does ShotMeta analyse images without uploading them?
ShotMeta runs Claude AI API on your local machine. Your images are converted to a smaller resolution and are analysed on your computer, not on a remote server. The tool reads each file, processes the visual content through the AI model, and outputs a complete metadata set. Nothing leaves your hard drive.
This matters for two reasons. First, privacy: if you shoot models, private property, or unreleased commercial work, uploading those files to a third-party server creates exposure you do not need. Second, speed: local processing removes the upload and download step, which on a 100-image batch with large jpg files adds up to meaningful time savings.
For a deeper look at how AI-generated metadata performs on approval rates, see our AI metadata myths post.
Step-by-step: how to keyword 100 photos with ShotMeta
Step 1: Download and install ShotMeta
Purchase ShotMeta from shop.drrave.com and follow the included installation instructions. It runs on Mac and Windows.
Step 2: Select your folder
Open ShotMeta, add your anthropic API (this is done once only) and select browse at the folder containing your processed images. Preferrably edited JPEG (RAW files are also supported).
Step 3: Run the analysis
Click Generate. The AI model analyses each image and writes a title, description, and keyword list. On a standard laptop, 100 images typically process in 7 to 12 minutes depending on file size and hardware.
Step 4: Export
ShotMeta embeds the metadata directly into the file headers using XMP/IPTC for all images and exports a CSV file for the video files. You can run a batch with both images and videos in the same folder.
One-time purchase, $29. Runs on your machine. No monthly fee, no cloud upload.
How long does keywording 100 photos actually take with ShotMeta?
Here is the realistic time breakdown for a 100-image batch:
| Task | Manual workflow | ShotMeta workflow |
|---|---|---|
| Title and description writing | 2 to 3 hours | Automated |
| Keyword generation (40 per image) | 2 to 4 hours | Automated |
| AI processing time | N/A | 7 to 12 minutes |
| Embed | Included above | Included above |
| Total | 4 to 7 hours | 7 to 12 minutes |
How does ShotMeta compare to other AI metadata tools?
Several tools now offer AI-powered stock metadata. The key differences are where processing happens, the pricing model, and what they export. For a full breakdown, see our roundup of the best AI metadata tools in 2026. The short version:
| Tool | Processing | Pricing | Export |
|---|---|---|---|
| ShotMeta | Local (no upload) | $29 one-time | CSV, XMP/IPTC embed |
| AutoMeta AI | Cloud upload | Subscription | CSV |
| GenMeta | Cloud upload | Credits/subscription | CSV |
| Picseta | Cloud upload | Free tier / paid | CSV |
If privacy and a one-time price are the priority, ShotMeta is the clear choice. If you are fine with cloud uploads and want a free tier to test first, Picseta is worth trying before committing.
Frequently asked questions
Does ShotMeta upload my photos to the internet?
No. ShotMeta runs AI locally on your machine. Your image files are never sent to a remote server. The analysis, title, description, and keyword generation all happen on your computer.
Which platforms does ShotMeta support?
ShotMeta generates metadata formatted for all including Adobe Stock, Shutterstock, Getty Images, and Pond5. It even sorts the keyword based on relevance which certain stock sites prefer like Adobe stock.
Can ShotMeta handle video files, not just photos?
Yes. ShotMeta analyses both photos and video clips and generates submission-ready metadata for each. This makes it useful for contributors who upload to both photo and footage sections of microstock platforms.
Will AI-generated metadata get my images rejected?
Platforms reject images for inaccurate or keyword-stuffed metadata, not for being AI-assisted. ShotMeta generates accurate, descriptive metadata based on what is in the image. For the full picture, see our post on AI metadata and stock photo approval myths.
Is ShotMeta a one-time purchase or a subscription?
One-time purchase at $29. No monthly fee, no per-image credit system. Note: you will need an Anthropic API which is a pay as you go service.
What format is the exported metadata in?
ShotMeta embeds metadata directly into image file headers using XMP/IPTC, so the data travels with the file in Lightroom and Bridge and exports as a CSV for video files, for upload to Deepmeta or Shutterstock.
Stop spending 6 hours on metadata. Do it in under an hour.
ShotMeta uses AI to write submission-ready titles, descriptions and keywords for your stock photos and videos. Runs locally on your machine. No cloud upload. One-time purchase.
Get ShotMeta for $29 →One-time purchase. Works on Mac and Windows.