Convert

Convert Document

Convert one document to another format — PDF ↔ Word, PDF → image, image → PDF and more. Synchronous, formatting-preserving, with an optional AI vision engine for hard scanned PDFs.

Free to try — no key required 40 credits / callAI vision: 200 creditsSynchronous

Overview

Upload a file and name a target_format; the source format is inferred from the uploaded filename’s extension. The response is a public R2 URL for the converted file. Like compression, this is on the free-try path.

POSThttps://api.bytefuse.in/api/v1/conversion/single

Request

Content type multipart/form-data:

filefilerequired

The document to convert. Must be non-empty; the source format is its filename extension.

target_formatstringrequired

Destination format — e.g. docx, pdf, png, jpg, txt.

quality_presetstringoptional

One of standard, premium, maximum.Default: standard

enginestringoptional

auto or vision. Any other value returns 422. vision only takes effect for PDF → DOCX.Default: auto

custom_settingsstringoptional

A JSON string of custom settings. Accepted but not consumed on this path.Default: null

Engines & quality

With engine=auto the API picks the right pipeline for your file:

Digital PDF → Word uses pdf2docx (quality ≈ 0.95).
Scanned PDF → Word is detected automatically (no text layer) and routed through OCR (ocr+python-docx, quality ≈ 0.88).
PDF → image renders at 150 DPI; a multi-page PDF returns a ZIP of per-page images.
Image → PDF and other pairs run through the conversion orchestrator.

When to use engine=vision

For complex or scanned PDF → Word where layout fidelity matters, engine=vision runs a premium AI pipeline (vision+python-docx). It applies only to PDF → DOCX; for other conversions the value is validated but ignored. It costs 200 credits instead of 40.

Response

200 OK, application/json:

200 OK

{
  "conversion_id": "b3c1e2a4-...-uuid",
  "status": "completed",
  "output_file_url": "https://<r2-host>/converted/report.docx",
  "quality_score": 0.95,
  "processing_time": 1.0,
  "engine_used": "pdf2docx",
  "metadata": {
    "source_format": "pdf",
    "target_format": "docx",
    "quality_score": 0.95,
    "processing_time": 1.0,
    "engine_used": "pdf2docx",
    "file_size": 48213
  },
  "created_at": "2026-07-04T12:00:00Z"
}

conversion_idstringoptional

Fresh UUID per call (not a persisted job id).

statusstringoptional

Always "completed" on success (synchronous).

output_file_urlstringoptional

Public R2/CDN URL of the converted file.

quality_scorefloatoptional

0–1 estimate of conversion fidelity.

processing_timefloatoptional

Seconds spent converting.

engine_usedstringoptional

The pipeline chosen — e.g. pdf2docx, ocr+python-docx, vision+python-docx, pymupdf-render, pymupdf-image.

metadataobjectoptional

Source/target formats, quality, timing, engine and output file_size.

created_atstringoptional

ISO-8601 UTC timestamp.

The response header X-Credits-Charged reports the credits billed.

Credits

Any conversion (default)convert.document40

engine=vision, PDF → DOCXconvert.pdf_to_docx.ai200

Examples

curl -X POST https://api.bytefuse.in/api/v1/conversion/single \
  -H "X-API-Key: dt_live_your_key" \
  -F "file=@report.pdf" \
  -F "target_format=docx" \
  -F "quality_preset=premium" \
  -F "engine=vision"

Errors

400Missing filename, or the uploaded file is empty.

422engine isn’t auto or vision, or the conversion engine couldn’t process the file.

402 / 429 / 503Insufficient credits, rate limited, or capacity/config error.

500Unexpected internal conversion error.

See Errors for the full error shape, and PDF to Excel for table extraction.

PreviousCompress PDF Next PDF to Excel