Skip to content
dialegein

Apache Tika

Java toolkit that detects and extracts text and metadata from over a thousand file formats through one interface.

Jobs Extract OCR

Operated by The Apache Software Foundation

Visit Apache Tika

Opens the vendor's own site.

Jobs covered

Open a task to see exactly what the vendor documentation supports.

Extract Browse all extract tools

detecting and extracting metadata and text from over a thousand different file types, including PPT, XLS and PDF, parsed through a single interface, in the Apache Tika toolkit

OCR Browse all ocr tools

optional OCR for images and PDF pages when an external OCR engine such as Tesseract is installed and configured

Where your files are processed

No server involved

A Java library, a command-line application and a server that you run yourself. There is no Apache-hosted document processing service, so a file parsed with Tika stays on your infrastructure. Note that 4.x can be configured to call vision-language-model APIs and remote OCR services; those are operator choices, not defaults, and they belong to the configuration rather than to the product. That is a property of the library itself: what an application built with it does with your file depends on that application, not on this.

A Java library and a server you run yourself, so documents are processed on your own infrastructure. Two documented exceptions to be aware of rather than surprised by: version 4 adds parsers that send content to Claude, Gemini or OpenAI for vision-language processing, and Tika can be configured to use remote OCR services.

Both are opt-in configurations rather than defaults, but a reader choosing Tika because it runs locally should know the path exists.

What it does

Tika reads a file and hands back its text and metadata, whatever the file is - PDF, Office documents, HTML, email, images, archives - behind one interface rather than one parser per format. It ships as a Java library, a command-line application, a REST server and a gRPC service, and it is the usual choice when a search index or a content pipeline has to digest documents it did not create.

Version 4 changed the default output to Markdown and moved parsing into separate processes, which is intended to isolate a parser crash from the host. It is isolation rather than a guarantee: resource exhaustion and other failures are not covered by it.

Version 4.1.0 is current, with 3.x still supported, and Java 17 is required.

Best for

Search platforms, archives and records systems that must ingest many formats without writing a parser for each, and Java teams who want one dependency for document text. It is the natural companion to a search engine rather than a tool for working on a single PDF.

Worth knowing

Breadth is the product, and depth is not: Tika's promise is that it handles over a thousand formats, not that its PDF text is as good as a dedicated PDF library, and a pipeline that depends on precise reading order or table structure should test it on real documents rather than assume. Version 4.0 is a major release with breaking changes, requires Java 17, moves parsing into forked processes and changes the default output to Markdown, so an upgrade is a project rather than a dependency bump.

The newer version also bundles parsers that call vision-language models - Claude, Gemini and OpenAI - which means a tool whose selling point is local document processing now has a documented path that sends your content to somebody else's API. And running the REST server exposes a document parser to a network: the project has had to tighten its defaults, and the 3.x line requires a flag named enableUnsecureFeatures before certain endpoints work at all.

Details, and where each comes from

Vendor statements, our own reading of their pages, and the things nobody has established yet, each answered separately. Where we have not confirmed something it says so rather than guessing, and a row nobody has established is greyed out.

Versions covered
Linux · Windows · macOS · Self-hosted
Website
tika.apache.org
Requires an account
Not applicable
Watermark on the free tier
Not verified
Maximum file size
Not verified
Batch processing
Yes
Data residency
Self-hosted
Open source
Yes
Self-hostable
Yes
Licence
Apache-2.0, with a number of bundled subcomponents under their own licences (documented in LICENSE.txt)

Apache-2.0, with a number of bundled subcomponents under their own licences (documented in LICENSE.txt)

Redaction
No redaction feature
Programmatic access
Embeddable library · Command line · HTTP API
Last availability check
Responded in 25 ms 21 October 2026

Checked automatically by our own runner, not reported by the vendor. 100% of our last 30 checks got a response.

What the redaction verdict rests on

From the vendor's own documentation, read 22 September 2026. Not tested by us.

“The Apache Tika™ toolkit detects and extracts metadata and text from over a thousand different file types (such as PPT, XLS, and PDF).”

What that does not establish

Recorded as `none` on absence: Tika reads documents and emits text and metadata, and nothing on the project's own page or in its release notes describes removing content from a file. It is a detector and a parser, not an editor.

What it does not establish: that the text it extracts is safe to share. Tika's whole purpose is to pull content out of a document, so the extracted text carries whatever was in the file - the opposite of redaction - and its metadata extraction will surface document properties that a redaction workflow is trying to remove.

https://tika.apache.org/

What the vendor says

Claims are labelled rather than repeated as fact. Where we have checked something ourselves, it says so and shows when. Checked means we read the vendor's own page and it says what we recorded - not that we tested the behaviour, unless the note says we did.

  • “The Apache Tika™ toolkit detects and extracts metadata and text from over a thousand different file types (such as PPT, XLS, and PDF). All of these file types can be parsed through a single interface, making Tika useful for search engine indexing, content analysis, translation, and much more.”

    Checked by us - the source says this Read 1 October 2026

    Confirmed. The quoted text appears in the rendered page as read on 2026-10-01.

    This claim was carried as reported rather than checked until this read.

    Standing: Not re-read against the live source

    Check it yourself: the source, the SHA-256, and how we checked
    Read
    https://tika.apache.org/
    From
    the text a browser renders
    SHA-256
    5920dc3af4d5784e2f4d13221a35f426df2218ac4bbfac50521782e6a92087db

    How we checked

    Read as rendered text in headless Chrome on 2026-10-01 and saved to the evidence archive, so the read can be repeated against the same file. Comparison ignores case, punctuation and hyphenation, which is why "open-source" on the record matches "open source" on the page.

    What this hash proves, and what it does not

  • “Apache Tika 4.0.0 has been released! The first stable release of the 4.x line switches the default output to Markdown, moves parsing into crash-isolated forked processes (tika-server, tika-app async/fork and tika-grpc all parse out-of-process via tika-pipes), adds vision-language-model parsers (Claude, Gemini, OpenAI) and in-process Tesseract OCR, and requires Java 17.”

    Checked by us - the source says this Read 1 October 2026

    Confirmed. The quoted text appears in the rendered page as read on 2026-10-01.

    This claim was carried as reported rather than checked until this read.

    Standing: Not re-read against the live source

    Check it yourself: the source, the SHA-256, and how we checked
    Read
    https://tika.apache.org/
    From
    the text a browser renders
    SHA-256
    5920dc3af4d5784e2f4d13221a35f426df2218ac4bbfac50521782e6a92087db

    How we checked

    Read as rendered text in headless Chrome on 2026-10-01 and saved to the evidence archive, so the read can be repeated against the same file. Comparison ignores case, punctuation and hyphenation, which is why "open-source" on the record matches "open source" on the page.

    What this hash proves, and what it does not

Each claim is quoted from the vendor's own pages. This section describes what we have and have not checked - it is not a judgement about anyone.

Sponsorship is distribution, not an endorsement or a promise of search rankings, AI citations, traffic, or sales. This listing is a free entry, compiled from published sources.

Visit Apache Tika Report a problem

Related reading

How this record was put together

Editorial review.
No editor has signed this record off. That is not the same as nothing having been checked, and the lines below say what has.
Source reading.
2 claims taken from the vendor's own pages, read between 1 October 2026 and 1 October 2026.
Website availability.
The vendor's site was last checked on 21 October 2026.
Behaviour testing.
We ran this ourselves. The file handling was watched directly.

Sources: tika.apache.org, read 2026-09-22 with headless Chrome, including the release history on the home page; the licence from the project's own LICENSE.txt. Project documentation, not independently tested; Tika was not installed and no document was parsed.