Output Formats¶

The search, lookup, cite, and canonical commands all share the same output controls: -f/--format chooses the shape, and -o/--output writes to a file instead of stdout.

Formats¶

Format	Flag	Best for	Available in
Text	`-f text` (default)	Reading in the terminal	all commands
JSON	`-f json`	Piping into other tools, batch downloads	all commands
BibTeX	`-f bibtex`	Reference managers, LaTeX	all commands
CSV	`-f csv`	Spreadsheets	`search` only

Text¶

The default. Human-readable, one paper per block, with title, authors, year, venue, and identifiers.

opencite search "graph neural networks"

JSON¶

Structured output, ideal as machine input. The JSON from a search is the natural input for batch downloads:

opencite search "graph neural networks" --max 50 -f json -o gnn.json
opencite batch-fetch --from-json gnn.json --convert -o ./papers

Add -v/--verbose to include extra fields such as per-source PDF locations, license, and version where the upstream APIs surface them.

BibTeX¶

Citation entries ready for your reference manager or LaTeX bibliography:

opencite search "graph neural networks" --max 20 -f bibtex -o gnn.bib

lookup can also append to an existing library in one step:

opencite lookup 10.1038/nature12345 --append-bib refs.bib

CSV¶

A flat, spreadsheet-friendly table. Available for search:

opencite search "graph neural networks" --max 100 -f csv -o gnn.csv

Writing to a file¶

Every format respects -o/--output:

opencite canonical "reinforcement learning" --min-citations 2000 -f bibtex -o rl.bib

Without -o, output goes to stdout, so you can pipe it:

opencite search "transformers" -f json | jq '.[].title'