First API call
Build your first Document AI app with a single prompt.
Made for coding agents. Less API research, more focus on product and outcome.
- Own EU infrastructure Your data remains in Europe.
- ISO 27001 active Security certified.
- SOC 2 Type II active International audit standards.
- 24+ years of experience Proven. Stable. Future-proof.
How would you like to start?
Enter a valid po_ut_ key to copy the zero-shot prompt.
Kept only in this browser tab. Pasting transfers the key to the coding tool.
Give PaperOffice to the coding agent
AI-readable PaperOffice API specification
https://api.paperoffice.ai/latest/docs/llms.txt Use with
- Cursor
- Claude Code
- Codex
- AI Coding Agents
Copy the instruction and paste it
Ready Document AI build instruction
Read in full:
https://api.paperoffice.ai/latest/docs/llms.txt
Task:
Build, run and verify a production-ready Document AI batch web app with PaperOffice AI-OCR.
Access:
Demo mode:
Use the preprocessed responses and Sandwich PDFs from this manifest: https://paperoffice.ai/demos/ai-ocr/batch/manifest.json
Clearly label the mode as a static demo. Never claim live processing.
Synthetic batch source files:
- https://paperoffice.ai/demos/ai-ocr/batch/sample1-source.jpg
- https://paperoffice.ai/demos/ai-ocr/batch/sample2-source.jpg
- https://paperoffice.ai/demos/ai-ocr/batch/sample3-source.jpg
Preprocessed demo artifacts:
- Responses: https://paperoffice.ai/demos/ai-ocr/batch/sample1-response.json | https://paperoffice.ai/demos/ai-ocr/batch/sample2-response.json | https://paperoffice.ai/demos/ai-ocr/batch/sample3-response.json
- Sandwich PDFs: https://paperoffice.ai/demos/ai-ocr/batch/sample1-searchable.pdf | https://paperoffice.ai/demos/ai-ocr/batch/sample2-searchable.pdf | https://paperoffice.ai/demos/ai-ocr/batch/sample3-searchable.pdf
- Complete package: https://paperoffice.ai/demos/ai-ocr/batch/paperoffice-aiocr-batch-demo.zip
Requirements:
- Download and process all three source JPGs together as one batch
- Run AI-OCR Complete with ocr_mode=complete, output_searchable_pdf=true, processing_lane=instant and client_wait=true
- Call the documented endpoint once per file; send every live upload as multipart/form-data in file_1
- Coordinate the three jobs as one batch and show per-file progress and partial results
- In demo mode, use only the preprocessed responses and PDFs without new API processing
- Fully implement upload, processing, HTTP 202 polling, errors and result
- For every file, render and offer downloads for the source JPG, OCR text, bounding boxes, response and Sandwich PDF
- Offer individual downloads and the complete demo batch download
- Use only endpoints documented in llms.txt
- Do not invent APIs, fields or download URLs
Result:
Implement the complete batch app, start it and verify all three source files.- 1 Paste into the coding agent
- 2 Run the generated app
Done!
Next station
Where to go next
The recommended next step in the developer funnel and two fitting branches.
