POST /jobs/upload

Upload files and get a cost preview. Credits are not deducted at this step.

Form fields

Field	Type	Required	Default	Description
files	File[]	Yes	—	One or more files (multipart/form-data)
ocr_model	string	No	premium	basic or premium
ocr_engine	string	No	easyocr	tesseract, easyocr, or rapidocr (premium only)
columns	string	No	defaults	Comma-separated or JSON array of column names
prompt	string	No	""	Extra instructions for the model
dpi	int	No	300	PDF rendering DPI
relative_paths	string	No	""	JSON array of relative paths for archive files
docling_refinement	bool	No	false	Enable LLM layout refinement for PDF/Office/HWP (3 md/page)

Request

curl -X POST https://your-domain.com/api/v1/jobs/upload \
  -H "X-API-Key: chu_live_xxxxxxxx" \
  -F "files=@document.pdf" \
  -F "ocr_model=premium" \
  -F "columns=날짜,계정과목,적요,입금액,출금액,잔액" \
  -F "dpi=300"

Response

{
  "job_id": "job-abc123",
  "status": "pending",
  "file_type": "pdf",
  "total_pages": 10,
  "total_files": 1,
  "media_duration_seconds": 0,
  "docling_refinement": false,
  "docling_refinement_pages": 0,
  "ocr_model": "premium",
  "ocr_engine": "easyocr",
  "has_media": false,
  "cost": {
    "pages": 10,
    "image_count": 0,
    "audio_seconds": 0,
    "video_seconds": 0,
    "ocr_model": "premium",
    "free_pages_used": 0,
    "points": 50,
    "usd": "$0.05"
  },
  "balance": 10000
}

Fields

Field	Type	Description
job_id	string	Job identifier — use for confirm/status/download
status	string	Always pending at this stage
file_type	string	pdf, docx, pptx, xlsx, hwp, hwpx, image, audio, video, archive, mixed (Markdown files report as mixed)
total_pages	int	Total PDF/Office/HWP pages detected
total_files	int	Total files (after archive extraction)
media_duration_seconds	int	Total audio/video duration
docling_refinement	bool	Whether docling refinement was requested
docling_refinement_pages	int	Pages subject to docling refinement (0 if disabled)
ocr_model	string	Model used (basic or premium)
ocr_engine	string	OCR engine used (tesseract, easyocr, or rapidocr)
has_media	bool	Whether audio/video files are included
cost.points	int	Credits that will be deducted on confirm (milli-USD)
cost.usd	string	Human-readable USD cost estimate
cost.ocr_model	string	Model used for cost calculation
cost.free_pages_used	int	Free pages applied (basic model only)
balance	int	Current credit balance in milli-USD (before deduction)

Office and HWP files

Office (.docx, .pptx, .xlsx) and HWP (.hwp, .hwpx) files are supported and routed through the Docling preprocessing pipeline (Office) or pyhwp (HWP). Page counts are estimated via Docling/pyhwp (defaults to 1 page if estimation fails).

curl -X POST https://your-domain.com/api/v1/jobs/upload \
  -H "X-API-Key: chu_live_xxxxxxxx" \
  -F "files=@document.docx" \
  -F "docling_refinement=true"

Markdown files

Markdown (.md) files are supported and processed at no cost (0 credits). The text content is used directly as the result without OCR/LLM processing. Markdown files are ideal when you already have structured content and want to convert it to CSV/XLSX/DOCX/PPTX via the download/convert endpoints.

curl -X POST https://your-domain.com/api/v1/jobs/upload \
  -H "X-API-Key: chu_live_xxxxxxxx" \
  -F "files=@notes.md"

Multiple files

Upload multiple files at once:

curl -X POST https://your-domain.com/api/v1/jobs/upload \
  -H "X-API-Key: chu_live_xxxxxxxx" \
  -F "files=@doc1.pdf" \
  -F "files=@doc2.pdf" \
  -F "ocr_model=premium"

Errors

Status	Meaning
400	No files, unsupported format, or missing filename
413	Total file size exceeds limit (default 200MB) or pages exceed limit (default 10,000)
502	File processing failed