ODS is an OpenDocument workbook, the native format of LibreOffice Calc. CSV is one table and nothing else, with no formulas and no cell formatting. LibreOffice sits between the two, and this page says exactly what happens to an ODS file on the way to becoming a CSV one.
What runs when an ODS file becomes CSV
LibreOffice 24.2, on a machine we rent and watch. Your ODS file is uploaded once, LibreOffice runs once, the CSV comes back, and neither file is kept. ODS to CSV is one of the 3649 pairs that engine was probed on with a real file, which is why it has a page here and why the pairs the probe could not prove do not.
What survives from the ODS into the CSV
Every value, as a value. What a formula worked out lands in the CSV as the number it produced, ready for anything that reads CSV.
A CSV that anything can open. No version, no schema, no library: this is the format you reach for when the thing on the other side is a script rather than a person.
What ODS to CSV costs you
Everything that is not the content itself: styling, headings, images, and any structure the ODS file used to describe itself. A CSV is one flat thing, on purpose.
Formulas, replaced by what they currently work out to. Measured on this pair: the SUM in the ODS file arrives as its answer and not as a SUM, so the CSV is a snapshot and not a model.
Every sheet but the first. LibreOffice writes one sheet into a CSV because a CSV has no way to hold two, and it takes the first one. A twelve-tab workbook comes back as one table, and the engine says so in a header rather than letting you find out later.
Where ODS and CSV files come from
ODS. LibreOffice Calc and OpenOffice write it natively. It is ISO/IEC 26300, which is why several public administrations ask for it by name in procurement, and why a file that arrives as ODS often arrives from one of them.
CSV. Everything writes it and nothing agrees on it. There is an RFC, 4180, and it describes common practice rather than commanding it: quoting, escaping and line endings all vary by producer. A CSV is the lowest common denominator, which is both why it is everywhere and why it surprises people.
If a model is going to read the CSV
Structure is what survives from the ODS file, and structure is what a model needs. ODS in, CSV out, with the heading tree intact instead of flattened into one long paragraph.
ODS to CSV, measured rather than promised
Running ODS to CSV against the real engine with a real file, LibreOffice wrote 89 bytes of CSV in 1.1 seconds, machine otherwise idle. That is one ODS file on one day and not an average, which is why the number is given together with the file that produced it.
The ODS to CSV verdict was reached by reading the internal type and finding a witness word, meaning the output was read back and recognised. Which check was used matters, because they do not all prove the same thing, and a status code of 200 proves nothing whatsoever about whether the CSV file has anything inside it.
The same ODS file, sent somewhere else
Other outputs the probe measured out of an ODS file, so the cost of choosing CSV can be read against something. Sizes do not compare across engines, because each family was probed with its own ODS witness file.
ODS to PDF: 12,884 bytes in 1.1 seconds, verified by a format signature.
ODS to HTML: 2,066 bytes in 1.0 seconds, verified by reading the internal type and finding a witness word.
Other ways into CSV, and what they measured
Among the published routes into CSV, ODS is the third largest output of the 5 measured. The witness files differ, so this ranks the probe run and not your document.
YAML to CSV: 39 bytes in under a tenth of a second.
XLSX to CSV: 89 bytes in 1.1 seconds.
JSON to CSV: 39 bytes in under a tenth of a second.
XLS to CSV: 89 bytes in 0.9 seconds.
What we will not pretend about ODS to CSV
An ODS file over 25 MB is refused before the upload finishes rather than after it, so you do not wait for a rejection.
An ODS to CSV run that passes 90 seconds is killed, and the LibreOffice process is killed with it. A run left behind would sit on one of the machine's two cores until somebody noticed.
LibreOffice is serialised on purpose, one document at a time per worker, because each soffice instance takes 150 to 300 MB on a machine with 3.8 GB. Measured ceiling: 23 ODS conversions in flight and roughly 0.9 a second sustained. Past that the queue answers 429 rather than falling over, which was tested at 32 at once.
Two thirds of the pairs you could imagine in this family do not exist, and it is worth knowing why before you go looking for one. LibreOffice converts inside whichever application opened your file: from a spreadsheet it writes spreadsheets, flat text, HTML and PDF, and it will not write a .docx no matter how it is asked. It returns nothing at all rather than something wrong. Of 113 combinations probed on this engine, 64 came back with a real file and the rest were refused outright.
This pair is one of the 64. It was proved by sending a real ODS file through the live route and reading the CSV that came back: not its extension, its internals, plus a marker word planted in the source and found again in the output. The traits named further up this page come from a second probe that looks for each of them one by one.
The engine tells the truth about the first-sheet limit rather than hiding it: the response carries an X-Sheets-Exported header saying so. It is mentioned here because a header is easy to miss and a missing sheet is not.
One ODS file at a time, chosen in the browser. There is nothing else to set up and nothing else on offer.
ODS to CSV: what people ask
What actually converts my ODS file to CSV?
LibreOffice does it, 24.2, on a machine we rent and watch. Not a browser trick and not somebody else service: the ODS file is uploaded once, LibreOffice runs once, the CSV comes back, and neither file is kept afterwards.
How long does ODS to CSV take?
On the file the probe used, LibreOffice took 1.1 seconds and wrote 89 bytes of CSV. That is one real measurement on one real ODS file, not an average and not a promise about yours: a larger ODS takes longer, and past 90 seconds the run is stopped.
What do I lose going from ODS to CSV?
The one to know about first: Everything that is not the content itself: styling, headings, images, and any structure the ODS file used to describe itself. A CSV is one flat thing, on purpose.
Why is my CSV so much smaller than the ODS?
Because a CSV holds the content and nothing else. Styling, fonts, images, headings and the file structure around them are all gone by design. Formulas are gone too, replaced by the values they had worked out at the moment of conversion.
My workbook had several sheets. Where are the others?
Only the first one is in the CSV, because a CSV is one table and has no way to hold a second. LibreOffice takes the first sheet. If you need them all, convert to XLSX instead, or run the conversion once per sheet from your own file.
Is ODS to CSV free?
There is a free allowance every month, and one ODS to CSV conversion costs half a credit against it. When the allowance runs out the tool says so and stops, rather than quietly handing you a worse CSV.