Initializing... drag & drop files here
Supports: ODG
Both ends of this conversion are vector formats, and this converter treats them that way. Your .odg is opened as a LibreOffice Draw document and written out through Draw's own SVG export filter (draw_svg_Export), so the shapes in the SVG are the shapes you drew: the same coordinates, the same curves, the same fills and strokes. Nothing is rendered to pixels at any point, and nothing is traced.
That distinction is the whole point of this page. Most routes into SVG are traces — a raster image goes in, an algorithm guesses outlines from colour boundaries, and a few thousand approximate paths come out. A Draw document never needs that, because the geometry is already sitting in the file waiting to be copied. Our pipeline names Draw's SVG filter explicitly rather than relying on LibreOffice's implicit pick, so an .odg always takes the vector route. The result is strictly better than anything a tracer could produce from the same drawing.
.odg onto the page or click "+ Add Files". OpenDocument Drawings are small — the whole document is zipped XML plus whatever media you placed in it — so uploads are quick even on a slow connection.image/svg+xml file that scales to any size without softening. Files are uploaded over an encrypted connection, processed on our servers, and deleted automatically after a few hours. No sign-up, no watermark.An OpenDocument Drawing is a package rather than an image: unzip one and you find content.xml, styles.xml, a manifest and a folder of embedded media. There is no picture inside to decode — the drawing exists only as a description of objects — so the conversion runs a real OpenDocument reader and then asks it to write the same objects out in SVG's vocabulary.
| Stage | What runs | What it produces |
|---|---|---|
| 1. Open the file | LibreOffice Draw's native OpenDocument import | A Draw document holding the original objects |
| 2. Export | Draw's SVG filter (draw_svg_Export) |
An image/svg+xml document built from those objects |
| Shapes and curves | Copied as geometry | Paths with the authored coordinates, not sampled outlines |
| Type | Vector | Never a raster of the letters |
| Placed photographs | Passed through | Raster data embedded inside the SVG file |
| Result | Resolution-independent | Scales to any size without blurring or stair-stepping |
Advanced Options on this page still carries a precision control inherited from the raster-tracing route that image sources take to SVG. A Draw document never reaches that tracer, so the control has nothing to act on. There is genuinely no setting to tune here, and that is the good outcome rather than a missing feature.
| Property | ODG | SVG |
|---|---|---|
| Full name | OpenDocument Drawing | Scalable Vector Graphics |
| Standard | OASIS OpenDocument Format; ODF 1.3 approved as an OASIS Standard on 27 April 2021 | W3C Recommendation; SVG 1.1 Second Edition, 16 August 2011 |
| Media type | application/vnd.oasis.opendocument.graphics |
image/svg+xml |
| Structure | Zip archive of XML parts plus embedded media | A single XML document |
| Pages | Several pages in one file | One canvas, no page model |
| Renders in a browser | No | Yes, natively, with no plugin |
| Edited with | LibreOffice Draw, Apache OpenOffice Draw | Inkscape, Illustrator, Figma, or a text editor |
| Typical use | Authoring and revising a diagram | Publishing it on the web or handing it to a designer |
SVG is the right output when the drawing is going onto a web page, into a design tool, or into a build pipeline that wants text-based assets it can diff and restyle with CSS. ODG to PDF is the better output when the drawing has several pages, when it is going to a printer, when it has to be filed or archived, or when the typography must render identically on a machine you do not control. Both routes run through the same LibreOffice Draw import and both keep the geometry as vectors — they differ in what the destination format is able to hold, not in how faithfully the source is read.
The actual shapes. Tracing is what you do when all you have left is pixels — a scan, a screenshot, a flattened logo — and it works by guessing where the edges were. Your .odg still contains the real objects, so the export copies them into SVG's own elements instead of reconstructing them. Nothing in this route rasterises the drawing, which means there is no stage at which precision could be lost and then approximately recovered.
No, and that is worth stating plainly because the page still shows one. The precision control under Advanced Options belongs to the tracing route that raster sources take to SVG, and a Draw document does not pass through it. Because the output is a copy of the geometry rather than a reconstruction of it, there is no accuracy dial to trade against file size — you already have the most accurate result the source can produce.
Yes. SVG has been a W3C Recommendation since long before the current browser generation — the 1.1 Second Edition Recommendation dates from 16 August 2011 — and every current desktop and mobile browser renders it natively. You can point an <img> tag at the file, drop it into an <object>, or paste the markup inline in your HTML and style parts of it with CSS.
The type is exported as vector content, so it is never a blurry raster of the letters. What SVG does not do is embed the typeface: it names the font and expects the viewer to have it. A machine that does not substitutes a metric-compatible face, which can change line breaks in a tight text box. If that matters, convert the drawing to PDF instead, or convert the text to curves in Draw before exporting so the letterforms are locked in as shapes.
Decide whether you actually want SVG. An SVG has a single root canvas and no page model, so a paged document does not map onto one file the way it maps onto a PDF. If you need every page, ODG to PDF is the direct answer. If you specifically need SVG, split the drawing in Draw so each file holds one page and convert them one at a time — that also gives you predictable, individually addressable assets, which is usually what an SVG is wanted for anyway.
Yes, and this is where a native export earns its keep. Because the objects come across as objects, you open the file and find shapes you can select, recolour and move. Traced output is the opposite experience: thousands of unnamed paths in stacking order, no groups, and colours quantised to whatever palette the tracer settled on. The same drawing exported natively stays workable.
They stay raster, because they always were. A photograph has no underlying geometry to preserve, so it is embedded in the SVG as image data and rides along with the vector content. The practical consequence is size: a diagram of boxes and arrows usually produces an SVG comparable to or smaller than the .odg, while a drawing built around large photographs produces a much bigger one. In that case convert to a raster format and let the format do what it is good at.
The geometry is not resampled, so the drawing itself arrives intact. What can change is anything SVG has no equivalent for: the page model, Draw's 3D rendering, and a handful of application-specific effects that have to be approximated with the primitives SVG offers. Fonts are the other one, since SVG references them by name rather than carrying them. None of that is the compounding, guess-the-edges kind of loss you get from a trace — it is the ordinary cost of moving between two formats with different feature sets.
Every upload travels over an encrypted connection, is processed on our servers, and is deleted automatically after a few hours along with anything it produced. Nothing is shared, published or reused, no account is required, and no watermark is applied. The ODG converter hub lists the other targets available for this source.