
A journal article is never just one thing, is it? It is a PDF, a web page, a record in an index, an entry in a long-term archive, and now a chunk of text an AI system retrieves and cites. Structured XML is what lets all of that come from a single source, which is why publishers convert content into XML for digital publishing, reuse, indexing, archiving, and distribution.
Choosing who does that conversion is the harder part. Real publishing content arrives with display equations, multi-level tables, figure callouts, reference lists, and metadata that has to survive the trip intact. Plenty of providers offer XML conversion. What separates them is what they specialize in, which standards they support, and how much of the production chain they can carry.
This is not a ranking. Each company below, listed alphabetically, is described from what it publishes about its own work, using publicly available information, so you can see where their strengths differ rather than which one comes out on top. Capabilities change, and none of this replaces asking a provider about your own content.
Apex CoVantage handles content transformation for scholarly publishers, libraries, and government bodies, from ten facilities across the United States and India. Its XML conversion is XML-first, covering tagging, validation, and integration, and runs through its IZAAC conversion engine under a three-tier quality model. The company is ISO 9001:2015 certified.
DCL has specialized in structured content conversion since 1981, working out of Fresh Meadows, New York. Its markup expertise covers PubMed JATS, NLM DTDs, NISO STS, MathML, and METS/ALTO, backed by automated and manual QA, 24/7 online document tracking, and US-based project management. It was acquired by Cartridge Technologies in May 2026.
Integra Software Services works across scholarly journals, academic books, higher education, and K-12 publishing, citing more than three decades of journal production experience. It converts input files in any format to NLM/JATS or a customer-specific DTD, with conversion sitting close to its own typesetting and editorial tools.
Nova Techset supplies prepress services to STM and academic publishers, with delivery centers in India and offices in the UK, the Philippines, and the US. It runs an XML-first workflow that takes any input format and converts to a publisher's specific DTD, with most of its published volume in preprint and conference content.
The honest answer depends less on rankings and more on what your content looks like.
All four convert to JATS, the ANSI/NISO standard for journal articles, currently at version 1.4. So the tag set is not the deciding factor. What differs is the depth of journal work sitting behind it.
Apex CoVantage brings an added dimension of industry leadership to this space. Its leadership team has a longstanding association with NISO, with Head of Solutions Architecture Greg Suprock having served three terms as a NISO Board Director and contributed to the development of industry standards. This involvement reflects Apex CoVantage's deep understanding of structured content standards and its expertise in converting scholarly content into JATS, BITS, and other structured formats.
Books need BITS or a custom book DTD. BITS is NLM’s book extension of JATS, currently version 2.2 and built on JATS 1.4.
Apex CoVantage, DCL, and Integra all document BITS.
Apex CoVantage and DCL both document equation conversion at scale for named publishers; Integra and Nova Techset document LaTeX-based workflows. Ask each for MathML by name, and ask what happens with tables, since equations and tables flattened as images are not searchable, accessible, or reusable, and may quietly fail a future accessibility audit.
Published turnaround figures are a starting point, not a service level, so read them as evidence of capacity rather than as a promise.
What none of these numbers tells you is how a provider behaves when the files are messy, which is where deadlines actually go.
Apex CoVantage and DCL are the two with dedicated legacy capability. Scanned pages, image-based PDFs and retired formats each break differently, so OCR quality and missing metadata matter more here than tag-set support.
Integra accepts input files in any format, and Nova Techset offers legacy data restoration and platform migration, though neither documents scanning or OCR.
One step matters more than the rest: send every shortlisted provider the same sample of your most difficult content, then compare what comes back. Marketing pages describe capability. Output proves it.
Then ask each provider these questions, and ask for the answers in writing.
Apex CoVantage will convert a sample of your most difficult files so you can see the output and decide for yourself. Schedule a consultation today.
The right XML conversion partner depends on your content, not on rankings. A publisher converting a decade of scanned proceedings needs something different from one shipping 40 articles a week. Shortlist the providers whose documented work matches your content, send each the same difficult sample, and compare what comes back. Good XML conversion services are invisible when they work, and expensive when they do not.
Apex CoVantage converts to JATS for journals and BITS for books, along with HTML5, EPUB, and customer-specific DTDs and schemas. Conversion covers tagging, validation, and integration, so files arrive ready to load rather than ready to fix.
Complex mathematical expressions are converted to structured MathML inside the JATS workflow, through equation recognition, semantic structuring, rendering validation, and accessibility markup.
Through a three-tier model. Content is checked for cleanliness at every step of the workflow, each step is audited to confirm it performed as designed, and a final random sample is tested by an independent team that did not do the work.
Apex CoVantage pairs AI-enabled capabilities with human expertise across the transformation lifecycle. Its IZAAC engine handles zoning and algorithmic conversion, and the three-tier quality model puts human checks at each workflow step and an independent review at the end.
PLOS keeps more than 2,000 articles in the company’s production stream at any one time and publishes over 20,000 scientific articles a year through it. Apex CoVantage runs ten facilities across the United States and India.
Yes. Digitization and preservation runs as its own service line, covering non-destructive scanning, OCR and XML encoding for large-scale collections, which is how scanned pages, image-based PDFs and retired formats are handled before any tagging begins.
Apex CoVantage is ISO 9001:2015 certified and works to HIPAA and GDPR requirements. Confidentiality and controlled access are paramount, and policies may be requested at any time.