Docling, as an API.
Convert PDFs, DOCX, PPTX, images, and more into Markdown or JSON.
No GPU required. No Python dependencies. Just works.
-H "Authorization: Bearer YOUR_API_KEY" \
-H "Content-Type: multipart/form-data" \
-F "file=@document.pdf" \
-F "output_format=markdown"
How It Works
Three simple steps to convert any document into structured data.
Speed Matters
Processing time for a 10-page complex PDF with CAD drawings
* Standard settings: OCR + Layout Analysis + JSON output. VLM-ready.
Why RunDocling?
Skip the infrastructure headaches. Focus on building your product.
| Feature | Local Docling | RunDocling API |
|---|---|---|
| Setup Time | Hours of dependencies | One API key |
| Infrastructure | Requires PyTorch & GPU | REST API |
| Performance | Slow CPU inference | GPU-accelerated |
| Maintenance | Memory leaks, updates | We handle it |
| Scaling | Manual, complex | Instant, automatic |
| Uptime | Your responsibility | 99.9% SLA |
Pricing
Start free, scale as you grow.
FAQ
What is RunDocling?
RunDocling is an independent managed API service that runs Docling for you. We handle all the infrastructure—CPUs, GPUs, and scaling—so you can convert documents without managing heavy dependencies or hardware.
Are you affiliated with the Docling team?
No. We are not affiliated with IBM or the Docling open-source project. We simply provide a managed hosting service that makes it easier to use Docling via a REST API.
How does it work under the hood?
We run docling-serve, the official open-source API wrapper for Docling. This means all parameters and options are fully compatible with the Docling library. We just handle the infrastructure so you don't have to.
Is my data secure?
Yes. Your documents are processed in isolated environments and automatically deleted after 24 hours. We do not store, share, or use your documents for any purpose other than conversion. Be sure to save your results before they expire.
Why is it faster than running locally?
We run Docling on dedicated GPU infrastructure optimized for document processing. No cold starts, no resource contention—just fast, real-time conversion without you managing any hardware.
What file formats are supported?
We support PDF, DOCX, PPTX, XLSX, HTML, Markdown, images (PNG, JPG, TIFF), and more. Output formats include Markdown and JSON with preserved tables, layouts, and metadata.
Why did you build this?
We're developers who got frustrated. Every time we needed to parse a PDF for a quick test or process documents in a pipeline, it was a mess—dependencies, GPU setup, memory issues. We just wanted something that works in 10 seconds. So we built it.