This tool demonstrates Stencila's format conversion capabilities. Upload a document or provide an arXiv, bioRxiv, medRxiv, or PMC identifier, choose an output format, and see how Stencila converts it.

Remote input supports arXiv IDs and URLs (best for preprints with HTML available), bioRxiv and medRxiv DOI identifiers and article URLs (only for preprints with full text XML available), and PubMed Central IDs and article URLs (only those in the Open Access Subset). Generic URL fetching is intentionally disabled for security reasons.

This public demonstration accepts uploads up to 25 MiB, allows five conversions per minute and 50 per day for each client IP, and times out conversions after 60 seconds. Concurrent requests share a small pool of containers, so demanding conversions may contend for CPU or memory and can slow down or fail during busy periods. Not every format supports every conversion option.

Files are kept only for the duration of a request. Uploads and file-based outputs use request-local temporary directories, and inline text outputs are returned from memory.

Normal concurrent requests use separate temporary directories. Filenames are sanitized, and public conversions disable persistent Stencila artifact caches.

The service returns a structured error. A container crash during a request could leave temporary files on ephemeral container disk until the container restarts, but they are not written to durable storage by this service.