You know the basics — add images, generate, export. But Meita.ai has a lot of settings and workflows that aren’t obvious at first glance. Here are 10 tips that power users rely on to move faster.
1. Use Banned Keywords to Clean Up Output
Every stock platform has keywords that get rejected or hurt your SEO. Go to Settings → Keywords → Banned Keywords and add a comma-separated list.
Common banned keywords to add:
- Generic terms:
image, photo, picture, illustration, stock - AI artifacts:
ai generated, artificial intelligence, midjourney - Platform-specific rejects (check your platform’s guidelines)
The AI will never include these in generated keywords, saving you from manual cleanup.
2. Set Target Title Length by Platform
Different platforms have different title length limits:
- Adobe Stock: Up to 200 characters
- Shutterstock: Up to 200 characters
- Freepik: Shorter titles perform better (80-120 chars)
Go to Settings → Output Rules → Target Title Length and set it for your primary platform. Enable Maximum Title Length to hard-cap titles that run long.
3. Process Selectively with Checkboxes
You don’t have to generate metadata for everything at once. Use the checkboxes in the table to select specific images, then click “Generate (N)” to process only the selected ones.
This is useful when:
- You want to use different prompts for different image types
- Some images failed and need re-processing
- You want to test settings on a small batch first
4. Use “Only Fill Empty Fields” for Partial Re-generation
Already have titles but need keywords? Or have keywords but no descriptions? Enable “Only Fill Empty Fields” in the Generate dialog.
The AI will skip fields that already have content and only generate missing data. This means you can:
- Write custom titles manually
- Run generation with “Only Fill Empty Fields”
- Get AI-generated keywords and descriptions while keeping your titles
5. Leverage Custom Prompts for Different Niches
The default prompt works great for general stock photography. But for specialized niches, custom prompts dramatically improve accuracy:
Food photography:
“Describe this as a professional food photograph. Focus on ingredients, cuisine type, presentation style, and dining context. Include keywords for food styling, recipe types, and dietary categories.”
Real estate:
“Describe this as a real estate or architecture photograph. Focus on room types, architectural features, design style, materials, and spatial qualities.”
Travel:
“Describe this as a travel photograph. Include the likely location, landmarks, cultural elements, time of day, and travel activities depicted.”
Keep “Automatically improve prompt with Meita” enabled — it blends your niche instructions with the optimized stock metadata template.
6. Use Drag & Drop for Faster Loading
Instead of clicking “Add Item” and navigating through the file picker every time, just drag files or entire folders directly into the app window. You can drop:
- Individual files
- Multiple files selected in your file explorer
- Screenshots and images copied to clipboard (Ctrl+V)
7. Configure Workers and Delay for Your API Plan
In Settings → Runner:
- Workers controls how many images process in parallel (default: 8). More workers = faster, but may hit API rate limits.
- Delay adds seconds between API calls. Set to 0 for paid API plans, 8+ seconds for Gemini’s free tier.
Recommended configurations: | Plan | Workers | Delay | |------|---------|-------| | Super AI (included tokens) | 8 | 1s | | Gemini Free | 2 | 8s | | Gemini Paid | 8 | 1s | | OpenAI | 5 | 1s |
8. Use Double Filter for Cleaner Keywords
Enable Settings → Runner → Double Filter to run keyword filtering twice after generation. The AI first generates keywords, then the filter removes:
- Duplicate or near-duplicate keywords
- Banned keywords
- Keywords that exceed your target count
The second pass catches edge cases the first pass might miss. It uses slightly more API quota but produces noticeably cleaner output.
9. Auto Export for Zero-Touch Workflows
In the Generate dialog, enable “Auto Export” and select an output folder. After generation completes for each image, Meita automatically exports the CSV and files — no need to manually click Export afterwards.
This is ideal for overnight batch processing: load your images, configure generation + auto export, and come back to ready-to-upload files.
10. Use Status Filters to Find Problems Fast
After processing a large batch, use the status filter pills at the top of the page:
- Click “Failed” to find and retry images that errored
- Click “Incomplete” to find images missing titles, descriptions, or tags
- Click “Defect” to review images the AI flagged for possible quality issues
- Click “Faces/Hands” to find images that may need model releases
Fix the flagged items, then re-generate only the selected ones. This is much faster than manually scrolling through hundreds of images.
Bonus: Keyboard Shortcuts
| Shortcut | Action |
|---|---|
| Ctrl+V | Paste image from clipboard |
| Drag & Drop | Add files from file explorer |
| Click image thumbnail | Open full-size preview |
| Arrow keys (in preview) | Navigate between images |
| Escape (in preview) | Close preview |
Want More?
- Getting Started Guide — First-time setup walkthrough
- Features Explained — Complete feature reference
- Got a question? Visit our Support page