Arelle is licensed under the Apache License, Version 2.0 (the “License”); you may not use this file except in compliance with the License.  You may obtain a copy of the License at  Unless required by applicable law or agreed to in writing, software distributed under the License is distributed on an “AS IS” BASIS, WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.  See the License for the specific language governing permissions and limitations under the License.


Source (Py3.2):
Click for GitHub Arelle Project GitHub project Arelle.
Mac OS App:
Mac OS (lxml)version - 2014-10-
filesize - 15.34 MB
Windows 64:
Windows x64 (lxml)version - 2014-10-11
filesize - 15.71 MB
Windows 32:
Windows 32 bit (x86 lxml)version - 2014-10-11
filesize - 15.09 MB
Linux x86 64: (Ubuntu, etc.)
Linux x86 64bitversion - 2014-10-11
filesize - 15.07 MB
Linux x86 64: (RedHat)
Redhat Linuxversion - 2014-10-11
filesize - 20.41 MB
Sparc Solaris: (10+ 64bit)
Sparc Solarisversion - 2014-06-09
filesize - 22.22 MB
Server (Py2.7): (App Engine)
Server (Python 2.7 source)version - 2014-10-11
filesize - 3.71 MB
XBRL database SEC (Postgres)
click for download instructions 2009-01 - 2014-09, 20GB pg_dump.gz

Copyright and Trademark

Arelle’s copyright and trademark are currently owned by Mark V Systems Limited (a California Corporation).


User’s Manual and Translating Arelle to your language, XBRL Spain, 2011 Manual de Usuario (Spanish), XBRL España, 2011 API Documentation.

Published Papers

Evolution and Future Trends for XBRL Development - Paper presented at 6th University of Kansas Conference on XBRL, April 25-27 2013, Lawrence, Kansas.
Enabling Comparability and Data Mining with the Arelle® Open Source Unified Model - Paper presented at First Conference on Financial Reporting in the 21st Century: Standards, Technology, and Tools in Macerata, Italy, 2011-09-09
Open Source & XBRL: the Arelle® Project – Paper presented at 5th University of Kansas Conference on XBRL, April 29-30 2011, Overland Park, Kansas.

Contact and support

Contact users group: Arelle-Users Group, leave a comment below, or on the blog.  You may also use e-mail to


Review and improvement of code, suggestions for functionality, documentation, and  just about everything is needed.  To contribute an Apache 2 Contributor License is required (modified to name this project).  If you are an individual, not under employee or works for hire agreement (e.g. student), please use the Individual Contributor agreement, otherwise the Corporate Contributor agreement. The reason why you need to use the corportate agreement if employed is because employment agreements and employment laws give intellectual property rights to the employer, so it is necessary for the employer to be part of that agreement.


62 Responses to Download

  1. i will evaluate this set of programas

  2. Xavi Balaguer says:

    Quiero evaluar el soft.

  3. GuoHui Chen says:

    Maybe, Python use more memory than other language like C#.

  4. Murray says:

    It loads okay, let’s see what it can do…

  5. Harold Kinds says:

    I downloaded the 16-8 version and the speed is tremendous!
    I downloaded the instance of MSFT in 4.3 sec.

    Unfortunately the save function seems not to be doing much…

    • admin says:

      The save function is only implemented for a few functions (such as for output formula results), perhaps it is best to remove from the toolbar (as there are right-mouse-button save options in pane windows that need save functions), until a more general pattern of use is identified.

  6. Amy says:

    Thanks, will take a look at it

  7. vineeth mohan says:

    I am not able to extract the linux build from the above link –
    The extractor is telling the file is corrupted.
    Please see what is happening.

  8. Ryosuke Fujita says:

    I downloaded source from github, and open as an eclipse pydev project. I executed but result is following. (My interpreter is python 3.2. OS is Windows vista 32bit)

    Traceback (most recent call last):
    File “C:\Users\fujita.ID\git\Arelle\arelle\”, line 19, in
    from arelle import XbrlConst
    File “C:\Users\fujita.ID\git\Arelle\arelle\”, line 1, in
    from arelle.ModelValue import qname
    File “C:\Users\fujita.ID\git\Arelle\arelle\”, line 139, in
    from arelle.ModelObject import ModelObject
    File “C:\Users\fujita.ID\git\Arelle\arelle\”, line 8, in
    from lxml import etree
    ImportError: No module named lxml

    I tried to resolve lxml dependence whole day long, but no progress right now.(tried to build & install lxml and so on)
    Are there documents which explains how to run on pydev?

  9. Jani Sillanpää says:

    I tried to install the source module arelle and I get error on row 191. Im using python 2.7.3 and setuptools-0.6c11:

    C:\Arelle> install
    Documentation production by Sphinx is not available: No module named sphinx.setu
    Traceback (most recent call last):
    File “C:\Arelle\”, line 191, in
    from cx_Freeze import setup, Executable
    ImportError: No module named cx_Freeze

    • Jani Sillanpää says:

      I see the arelle source is Py3.2. Ok that might explain why it doesnt work on 2.7.3. Ill try installing 3.2.

      • Jani Sillanpää says:

        I installed python 3.2 but now I get error regarding lack of
        setuptools. However I cant install setuptools since its not compatible
        with python 3.2?

        C:\Arelle> install
        Documentation production by Sphinx is not available: No module named sphinx.setu
        Traceback (most recent call last):
        File “C:\Arelle\”, line 190, in
        from setuptools import find_packages
        ImportError: No module named setuptools

        • admin says:

          I found it easier to use the active state builds, they had setuptools more available (but still painful to configure). Keep in mind that Python 3.3 is almost out, and we will probably switch when it is released (August plans). Python 3.3 has some unicode support that is really needed, may get the regex package that is needed for european applications, and has important performance improvements (such as the decimals package).

      • admin says:

        see prior response, both 2.7.3 and 3.2 work, but need significant investment in script writing and tailoring

    • admin says:

      The is configured for development builds. The 2.7.3 mode is intended to produce automatic documentation. The 3.x mode is intended to work with various scripts such as buildWinDists.bat for Windows, for MacOS, and for Ubuntu. In all cases getting to build a distribution and work with cx_freeze or platform installers is a lot of difficult work. I suggest studying the pre-built scripts to make your own

  10. Mauricio Ahumada says:

    I’ve tried de software but found several problems. For instance, the validación system seems to be broken, because i’ve loaded an valid XBRL (validated with Fujitsu Instance Creator, and UB Matrix processor engine) but when validated with Arelle, a LOT of error shows up saying that some facts are not dimensionally valid, also the errors do not seems the be the same everytime… We were looking for an alternative to our current validator and Arelle looked promising, but we cannot use it with the current inestability. Greetings.

    • admin says:

      Dear Mauricio,

      Please either (a) create an issue report on, or (b) e-mail to We need to know (1) what platform you run on (OS, Memory), (2) Arelle version and date (from help->about), (3) what kind of validation you are doing (formula, calculation, 2.1, etc), and (4) provide a zip file of an instance and its DTS that show this instability. We take this seriously and need sufficient information to clarify the issue.

  11. Michael Mason says:

    Is it possible to get Arelle to work on 32-bit linux? We have a development machine that is a bit older and I’d like to install Arelle on it.

    If I install Python 3.2 on Ubuntu 32-bit machine and build from source, should it work?

    • admin says:

      It should be no problem to just run Arelle (from sources) on 32 bit linux, but to do builds (of the installable product) takes a lot of preparation and work for ubuntu. Arelle can also be run (in server mode,, not GUI) from the 2.7 zip.

      To do builds, there’s a lot of messing around to get lxml, a few other modules needed, rebuild cx_freeze from scratch for 3.2, and adapt the shell script (in gitHub).

      Keep in mind that Python 3.3 is just about released, and will have some important unicode string handling and decimals features that Arelle will take advantage of (as soon as it’s officially released), and all the preparation work will have to be repeated for 3.3.

  12. Chao Li says:

    Seems can’t use it to open the taxonomy, always hit an error: “the target machine actively refused it” when the software try to reach files from, etc. While on the same computer, other XBRL tools works fine.

    Can you pls help out? many thanks.

    • admin says:

      Maybe this should be handled by a private e-mail to, or if you find a bug, enter to Please indicate (or provide in the e-mail) what you are trying to open (such as by attaching or referencing the instance and taxonomy). Please indicate if your configuration includes a firewall or proxy server, and if so, what the platform is, the type of proxy server if any, and whether you tried the various proxy setup configurations (automatic or manually entered by the tools, internet, proxy setup dialog). That will allow us to help you further.

  13. Jake says:

    I installed the latest version for Windows 64 (and 32) and get the following error when I try to start the web server. GUI works fine. I don’t have any experience in Python and have no idea what it means.

    C:\Program Files\Arelle>arelleCmdLine.exe –webserver localhost:8080
    Traceback (most recent call last):
    File “c:\python32x86\lib\site-packages\cx_Freeze\initscripts\”, line 27, in
    File “”, line 15, in
    File “C:\Users\Herm Fischer\Documents\mvsl\projects\Arelle\ArelleProject\src\arelle\”, line 38, in main
    File “C:\Users\Herm Fischer\Documents\mvsl\projects\Arelle\ArelleProject\src\arelle\”, line 236, in parseAndRun
    NameError: global name ‘logger’ is not defined

    The old version of Arelle web server used to work without issues too but now I cant install the old version as there doesn’t seem to be a link to download old versions

    • admin says:

      We’ll come up with a way of browsing old versions (they’re in a wordpress directory layout). Do you still get this error in the current version?

  14. Vladimir says:

    Downloaded both version for Windows x64 and Linux x64. Windows version runs ok, both GUI and server. But on Linux environment I’m getting:

    [user@linux arelle]$ ./
    ./arelleCmdLine: /usr/lib64/ no version information available (required by /media/persistent0/arelle/
    ./arelleCmdLine: /lib64/ version `GLIBC_2.15' not found (required by /media/persistent0/arelle/
    ./arelleCmdLine: /lib64/ version `GLIBC_2.14' not found (required by /media/persistent0/arelle/
    ./arelleCmdLine: /usr/lib64/ no version information available (required by /media/persistent0/arelle/

    Running on Amazon ALM Medium instance: Linux arellehost 3.2.34-55.46.amzn1.x86_64, 4096 Mb memory, Python 3.3

    • admin says:

      The pre-built linux version has been tested only on Ubuntu, not Amazon ALM, this developer does not have access to Amazon ALM.

      The Python versions supported are 3.2 and 2.7, we anticipate some development work needed for Python 3.3 (and are waiting for 3.3 distributed by ActiveState).

      Some users have been successful, in their organizations, with Amazon ALM, working with Arelle source code from gitHub, but have done that within their company projects. From the trace above, one could infer that some libraries seem to be missing compatible versions.

  15. Jarkko R says:

    Thanks for great application!
    I have used it for a while to browse taxonomies and viewing xbrl instance documents. However, I updated my installation and this new version does not open taxonomy (frow web) any more; it is loaded to memory but no presentation or concepts are presented in Arelle anymore. I tried this with a few older patch/version available in Archive as well with same results. However, loading a taxonomy (without instance file) works great for example in version published 2012-08-04. Is this a feature or bug ?

    • admin says:

      This should be working, but need more information to debug. Maybe a private message to with the example to test on. Also you may try clearing the cache (it may have an html file somewhere that it expects an xsd or xml file), tools->internet->clear cache.

  16. Dave Dansing says:

    I cannot open the linux download. I get the following error when I try to unzip it:

    gzip: stdin: not in gzip format
    tar: Child returned status 1
    tar: Error is not recoverable: exiting now

  17. Nils Wilhelm says:


    i just wanted to test arelle on linux and downloaded the current version. Using “tar -xzvf arelle…” i get the following error:

    “gzip: stdin: not in gzip format
    tar: Child returned status 1″

    Did i miss something or is there a problem with the archive?

    Best regards and thanks in advance

  18. I have been using Arelle for quite sometime, hvae never faced the following error

    [xmlSchema:valueError] Element attribute substitutionGroup value error: xbrli:item, Namespace is not defined by an import or include element – IN-GAAP/in-gaap-2012-03-31.xsd 96

    could you explain what it means and how to resolve
    I am connected to internet, i reinstalled the application

    • admin says:

      This is a schema validation message, it could be that the referenced substitutionGroup value has a qname whose corresponding namespace is missing an import, anyway, please e-mail sample files to so we can verify this.

  19. Jay says:

    Hi, I was wondering if there is a way to get a version of Arelle that works on python 2.5.x


    • admin says:

      No way for Python 2.5.x. The current code base supports Python 3.2. (3.3 is likely to be skipped but 3.4 is likely to be supported in the future). There is also limited support for server-level use of Python 2.7 (such as for GAE and command line, but not GUI). 2.5 lacks too many features that are critical to Arelle. Please consider a platform that is “current”, e.g., when Python 3.4 is then-current version in the future, after a while it may be required. You can always “build” Python 3.2 from scratch (but that takes a lot of detailed system and language knowledge).

      • Jay says:

        Hi, thanks for replying.

        I am running Arelle using Python 3.3 and all is working apart from the GUI.

        On Unix, when I try and launch the GUI for Arelle, I get the following error:

        import _tkinter # If this fails your Python may not be configured for Tk
        ImportError: No module named ‘_tkinter’

        • hank says:

          Me too. I have ubuntu 14.04 and python3.4 is the default python. But in arelle the python is 3.3, I can run “import _tkinter” with no errors in the default python3.4

          • admin says:

            most of arelle seems to work on 3.4, I had to change the logging API to work with 2.7 – 3.4, if you find specific things broken in 3.4 it would be great to start a Jira ticket on 3.4 glitches (

  20. Jay says:

    Hi, sorry for re posting this here, I think I posted it in the wrong section before.

    I wanted to find out, how does arelle distinguish what the presentation taxonomies are when dealing with COREP taxonomies.

    COREP taxonomies are split into template,dimension,primary Arelle can automatically load the primary taxonomy for a given template taxonomy in its “presentation” tab, when opening a DTS.

    How does this work?

    • admin says:

      I’m not sure how to answer your question, XBRL 2.1 uses “discovery” from the entry point (xsd or instance) to find and build up an object model of concepts, relationships, and such. This is a built-in part of XBRL. It’s quite lengthy to explain, a very technical description is in the XBRL 2.1 Specification.

  21. Jay says:


    also is it possible for Arelle to render XBRL instace?


    • admin says:

      Arelle can render XBRL instances in several manners, depending on the contents, type of rendering, and many other unknowns. In your other postings you referred to COREP, and if it is a current COREP instance using table linkbases, these are rendered, or if pre-table linkbase COREP, then the facts can be rendered in a simple columnar style.

      • Jay says:

        Thank you for your response, would you know where/if at all any samples of COREP instances with table linkbase available?

        Thanks again for your help.

        • Mark says:

          First thanks so much for producing this software, this whole XBRL thing is frankly daunting and you’re the first ray of hope I’ve found this week.

          For the poster above there is now a draft COREP taxonomy here :

          I can open this up in the Arelle GUI and create an instance file. If I can figure out how to navigate the schema and make something which looks like the COREP excel templates I’m home free.

          Is there any way to pull elements from the schema and design an XBRL instance? I can’t quite figure that out.


          • admin says:

            There are several options for this, (1) the concepts and linkbases can be viewed in the GUI and output to many formats of files (copy to clipboard from GUI or (2) using the command line or web services to save the concepts or linkbases to various file formats (xml, json, csv), or (3) outputing a “Loadable Excel” worksheet (by plugin saveLoadableExcel) which can be used as a template to prepare instance data (the loadFromExcel plugin presently can generate an extension taxonomy, and will be enhanced to also save instance documents), or (4) something cobbled out of the xbrlDB plugin which produces the XBRL-US public postgres database, Rexter/Titan/Cassandra graph database, or RDF turtle output.

  22. Jay says:


    I wanted to find out if Arelle currently supports

    XBRL Validation Message 1.0


  23. Hi! I have been trying Arelle as a web server on Windows and it works fine. Now, I am trying to make it work on Linux as a web server as well, but none of the installers seems to work for me. The 64-bit version seems to require Python 3.2, and I do not have it installed on my server. Then, I downloaded the server version (perfect for my Python 2.7) but the start script seems to look for an arelleCmdLine executable file that does not exist, but there is an file so, I decided to try this:

    # python –webserver localhost:8080

    It results in an “Invalid arguments” error. I also tried copying the arelleCmdLine file from the 64-bit installer and I get errors related to Python’s version. I just gave me up with the 64-bit version.

    What is the right way to start the web server from the server version? Also, is it possible to run Arelle from sources on Python 2.7?

    • admin says:

      Please contact The binary builds for Linux do not need python installed on your server, and should support web server operation, either using the built-in web server or using Arelle as a cgi service by the web server. We’ll need to know what kind of Linux you use, we do separate builds for Linux, RHEL, Solaris, and MacOS.

  24. Seb says:


    i’m getting exceptions after opening Open Tables defined in and

    [arelle:resolutionException] Exception in resolution of definition node table:aspectNode: ‘FactPrototype’ object has no attribute ‘isItem’ – 45
    Traceback (most recent call last):
    File “C:\Users\Herm Fischer\Documents\mvsl\projects\Arelle\ArelleProject\src\arelle\”, line 387, in expandDefinition

    To reproduce (tested in Arelle x64 2014-02-02):
    Load the taxonomie and open table: S.06.02.01

    I’m looking forward for a fix, i hope it’s no big deal!

    Thanks alot!

    Best Regards


  25. Francesco says:

    Hello, we are able to open the eba Taxonomy… but how can we open a COREP xbrl instance? Our regulatory software already produces the xbrl instance document and we want to check if the file uploaded on the regulator’s website is valid and reflects the correct figures.

    I don’t think is as straight forward as opening an .xbrl file extension? is it?

    here the errors that we are receiving:

    retrieving %(URL)s
    [FileNotLoadable] File can not be loaded: – EBA LIQREP_LCR_SOLO.xbrl 9
    [xbrl:schemaImportMissing] Instance facts missing schema definition: eba_met:mi129, eba_met:mi277, eba_met:mi287, eba_met:mi323, eba_met:mi324, eba_met:mi325, eba_met:mi53, find:fIndicators – EBA LIQREP_LCR_SOLO SSBEL v3_10_30_2014.xbrl 29628, 29631, 29634, 29637, 29640, 29643, 29644, 29645, 29646, 29647, 29648, 29649, etc. etc.
    loaded in 23.09 secs
    no relationships for
    no relationships for Presentation
    no relationships for Calculation
    no relationships for Dimension
    views 0.00 secs

    • admin says:

      in some cases it may be helpful to install a taxonomy package (which has the corresponding EBA distribution), in the GUI help->manage packages, install locally. It can also be disabled if the package is redirecting URLs that shouldn’t be redirected when working with other projects

      validation with all the formulas from previous EBA versions hasn’t been profiled and optimized yet, on the plate of near term tasks

      • Adam says:

        I am using Arelle to process iXBRL (.html) and XBRL (.xml) files. I use Arelle from the command line, to create various view files.

        Thank you for your wonderful software.
        I have two questions:

        (a) Arelle works well with the html files, but I get the following error with all xml files: “no relationships for Dimension”. I also cannot create a dim view from these files.

        Do you think your advice to Francesco would apply for me also? I can send you example xbrl files if you like.

        (b) Processing each file takes a few seconds, and I notice that the executable is accessing the internet when parsing each file, despite the fact that it is caching many files locally. Is there any way that I can speed up execution somehow?

        Thank you

        • Adam says:

          Please disregard (a). In the dataset I’m using, I think that the .html files all use a dimensional taxonomy, but the .xml files all happen to use a non-dimensional taxonomy. Hence, the message from Arelle is not an error message, just a friendly informational message.

          If you have advice regarding (b) it would be appreciated.

  26. Adam says:

    The command line options to save facts and factTable don’t seem to work. Also, the GUI method to save the Fact List just generates an error

  27. Chris Brown says:

    I have been running Arelle version issued 14th of May 2014 on Windows 7 64bit with 8GB RAM with Intel i5 processor for the past weeks. Loading and validating a simple XBRL file has been taking proximately 240 sec. but when I process the same XBRL on a Windows Server 2008 R2 with the same amount of RAM and a similar processor it takes twice the time. Any successions what can be the cause for this?

    • admin says:

      It could be the CPU processing difference between the chipsets, or anything else. It might help to check some comparison of single-core performance between the machines.

Leave a Reply

Your email address will not be published. Required fields are marked *

You may use these HTML tags and attributes: <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <strike> <strong>