Transpic AI - Private AI image translator that translates photo text while preserving original layout

Core Features

Transpic AI processes uploaded JPG, PNG, and static WebP files without publishing them as permanent public assets. All uploads use private storage and short-lived access URLs, with no permanent public hosting of user source images. The layout-aware text detection system identifies distinct text regions and their positions on the source image, so translated text blocks stay attached to the correct sign, speech bubble, price label, or table cell instead of being output as a separate plain text file. After automatic translation completes, users open the built-in editor to correct wording, resize text blocks, adjust font color, or reposition text before final export. Output files preserve the exact dimensions of the original uploaded image. The service supports 60+ target languages covering major Asian, European, Middle Eastern, and African languages for multi-region localization workflows.

Supported Use Cases

General mode

Suitable for posters, restaurant menus, app screenshots, street signs, learning material, and everyday photos with mixed text placement.

E-commerce mode

Translates product headlines, feature callouts, prices, and promotional copy while keeping product photos, color palettes, and visual purchase hierarchy fully intact. This eliminates the need to rebuild entire product posters from a blank canvas.

Manga mode

Optimized for short dialogue text in speech bubbles and panel-based reading flow. Translated text fits existing bubble positions to preserve original manga reading structure.

Document screenshot mode

Translates tables, forms, reports, and scanned page images while retaining rows, columns, labels, and numerical contextual relationships. Direct PDF and Word uploads are not currently supported.

Step-by-Step Workflow

1. Upload your image: Drop a valid JPG, PNG, or static WebP file into the browser uploader. The server validates file type, size, dimensions, and animation status before starting processing.

2. Select language and mode: Pick your target language and choose the workflow mode that matches your image type: general image, product graphic, manga page, or document screenshot.

3. Translate image text: The system detects all text regions, runs translation using selected AI models (Gemini, DeepSeek, ChatGPT, or Grok), and generates positioned text blocks, showing real-time progress for each job.

4. Compare, edit, and export: Use the interactive side-by-side comparison slider to check results. Adjust wording, font size, or text placement as needed, then export the final translated image.

Pricing Plans

All paid plans are billed annually and include full credit allowance upfront. Every plan provides watermark-free export and full commercial use rights.

• Free plan: 30 signup credits plus 10 daily credits that expire at 00 UTC. Supports up to 3 images per batch, 1 target language per job. Includes full editor access, all translation modes, and 2 free OCR extractions per day.

• Starter plan: $9.90 per month billed annually at $99 total. 60,000 total credits, up to 20 images per batch, 2 target languages per job, maximum 15MB image size, 30 days of editable result history, 500 free OCR images per month.

• Pro plan: $19.90 per month billed annually at $199 total. 144,000 total credits, up to 50 images per batch, 5 target languages per job, maximum 25MB image size, 90 days of editable result history, 1500 free OCR images per month. Users can run High quality mode that supports cross-page context.

• Studio plan: $79.90 per month billed annually at $799 total. 600,000 total credits, up to 100 images per batch, 10 target languages per job, maximum 50MB image size, 180 days of editable result history, 5000 free OCR images per month. Credit costs: Fast/Balanced translation consumes 10 credits per result. High quality translation consumes 20 credits per result on Pro and Studio plans. One result equals one image translated into one target language. Text removal and OCR operations also consume credits from the same balance. Failed processing jobs release reserved credits automatically and do not consume user allowances.

Known Limitations

The service does not support direct PDF or Word file uploads. Images with very small lettering, curved display type, heavy texture, or overlapping artwork may require manual wording, size, or placement adjustments in the editor after automatic processing.