GLPI 12 is the latest major release of the open-source ITSM solution. It introduces a completely redesigned Knowledge Base, gives administrators greater control over active user sessions, and ensures that internal deadlines automatically follow the team responsible for the work. The release also strengthens platform security, with no additional configuration required.
In this article, we explore the key improvements GLPI 12 brings to users and IT teams, as well as the essential points to consider before migrating.
As part of GLPI's release cycle, each major version introduces new features, enhancements, and updated technical requirements. GLPI 12 follows GLPI 11 and marks the end of support for GLPI 10.
The guiding principle behind this release is simple: bring GLPI even closer to the way teams actually work. A ticket moves from one team to another. A procedure is difficult to find. A user session remains active following an incident. GLPI 12 provides a built-in response to each of these everyday situations.
The comprehensive GLPI 12 changelog provides a detailed overview of every change. Below, we highlight the key improvements you need to know about.
This is one of the most significant changes introduced in GLPI 12. The Knowledge Base interface has been completely rebuilt to provide a smoother and more intuitive experience. Reading and editing now take place within the same interface, eliminating the need to navigate between multiple tabs and making knowledge easier to access, manage, and maintain.
In practice: a modern editor with quick commands, an always-visible tree with favorites and drag-and-drop functionality, public article sharing, comments anchored to specific passages, version comparison, and translation directly within edit mode.
Categories are no longer used as a separate structure. Articles can now directly contain other articles, making content organization simpler and more intuitive. When upgrading, existing categories are automatically converted into articles, and their visibility settings will need to be configured again.
Documents attached to tickets can be hidden from requesters, while users who are replaced on a ticket are automatically notified. Ticket search criteria can also be restricted by profile, ensuring that each user group only sees the information relevant to them.
Project tasks now include reminders, while forms allow multiple items to be added within a single question.
Benefit from new filters, including entity, multiple groups or statuses, SLA, technician, and operating system, as well as new dashboard panels displaying tickets by group and status, tickets by category and entity, and computers by age. Dashboard performance has also been improved for faster display and a smoother user experience.
An OLA (Operational Level Agreement) defines the internal deadline commitments between IT teams. Until now, a ticket carried a single OLA timer. When a ticket moved from the network team to the workstation team and then to the applications team, the accumulated delay could ultimately fall on the last team involved.
With GLPI 12, each team involved in resolving a ticket has its own deadline and its own late status. This provides service managers with greater visibility into where time is being spent or delays occur, while ensuring that each team is measured against the commitments it actually controls. The result: greater accountability, more accurate performance tracking, and better visibility across IT service delivery.
GLPI 12 keeps track of every active session, giving administrators clear visibility into who is currently signed in and allowing them to terminate any session instantly with a single click.
This is the answer for situations where changing a password is no longer enough: a lost or stolen laptop, a password exposed in a data breach, or a compromised integration. Until an active session is revoked, it remains valid.
Some sensitive administrative actions now require additional identity confirmation, even when you are already signed in. Simply verify your identity using a two-factor authentication code, with your password available as a fallback. For organizations requiring separation of duties, this security principle is built directly into the product, helping strengthen access control and protect critical administrative actions.
Requests from the GLPI server to the internal network are blocked by default, unless explicitly authorized, helping to prevent potential pivot attacks. When enabling inventory, users are now prompted to select an authentication method. It is still possible to proceed without authentication, but this option must be explicitly selected and is clearly flagged by GLPI as not recommended. The underlying technical foundation has also been strengthened, with consistent protection applied across all database access.
Security and privacy settings are now easier to access and manage. A dedicated Security menu brings together password policies and two-factor authentication (2FA), while the Data and Privacy menu centralizes data retention and purge rules. For GDPR management, everything you need is now conveniently available in one place.
API. GLPI 12 introduces version 3 of the High-Level API, running alongside previous versions. Your existing integrations remain fully operational, giving you the flexibility to migrate to the new API when you are ready.
Accessibility. GLPI 12 takes an important first step toward meeting RGAA and RAWeb accessibility criteria, with improvements to keyboard navigation, page structure, contrast, and screen reader compatibility. This work will continue across future releases. For organizations responding to public sector and healthcare tenders, these improvements provide concrete support for accessibility requirements.
Documentation. The GLPI user documentation has been completely rewritten to provide a clearer and more accessible experience. It now includes a Quick Start section, interactive demos, and comprehensive documentation in French, helping users get started and make the most of GLPI more easily: help.glpi-project.org.
A major release may introduce changes to existing configurations. Before upgrading to GLPI 12, we recommend checking the following eight key points to ensure a smooth migration.
Plugin developers can find further technical details in the relevant GLPI 12 developer documentation.
GLPI is open-source software and can be used without a license. A GLPI Network subscription, whether deployed on your own infrastructure or through GLPI Network Cloud, provides direct support and expertise from the official publisher. This can be complemented by the expertise of an integration partner, who can support your migration project, including carrying out a pre-migration audit to help ensure a smooth and secure transition to GLPI 12.
Plugins must be adapted to ensure compatibility with GLPI 12. GLPI Network plugins are ported to the new version. For community or in-house plugins, we recommend checking compatibility with their respective maintainers and testing them before migrating.
Your existing knowledge base categories become articles, with their previous content converted into sub-articles. They remain invisible until an administrator assigns the appropriate access rights, which are then inherited by their content.
Yes. Previous API versions remain available and can continue to be used. API v3 is only used by new integrations that specifically request it.
Following the release of GLPI 12, GLPI 10 is no longer supported, except for customers who have subscribed to GLPI Network Extended Support. Organizations still running GLPI 10 should therefore plan their migration to GLPI 11 or GLPI 12.
In the comprehensive GLPI 12 changelog and on the GitHub releases page.
GLPI 12 modernizes the knowledge base, gives administrators greater control over user sessions, ensures internal deadlines follow the teams responsible for them, and strengthens security by default. It also introduces a new API version, initial accessibility improvements, and fully revised documentation.
To prepare for migration, eight key points should be reviewed, with particular attention to knowledge base visibility and plugin compatibility. With a dedicated test environment and, when needed, the support of a GLPI Network partner, migrating to GLPI 12 can be planned and managed with confidence, following the best practices of any major upgrade project.
GLPI-Agent 1.20 has been released. You can download it on the GLPI Agent github project: https://github.com/glpi-project/glpi-agent/releases/tag/1.20
You can check changes details in the official online Changelog available here: https://github.com/glpi-project/glpi-agent/blob/1.20/Changes
This new version fixes the following security issues:
.xml file overwrite if local_store is enabledWe strongly encourage you to update your agents.
The release also includes some bug fixes and enhancements. Here are the more important ones:
About packaging, here is what you should retain:
We strongly encourage you to update your agents to ensure optimal security and operational performance.
You can download it on the GLPI Agent github project: https://github.com/glpi-project/glpi-agent/releases/tag/1.18
This new version is essentially a security fix release for the following security issues:
The release also includes few bug fixes:
You can check changes details in the official online Changelog available here: https://github.com/glpi-project/glpi-agent/blob/1.18/Changes
About packaging, here is what you should retain:
This time, we strongly encourage you to update your agents.
You can download it on the GLPI Agent github project: https://github.com/glpi-project/glpi-agent/releases/tag/1.17
This new version comes with some fixes and enhancements and here are the essential ones:
You can check changes details in the official online Changelog available here: https://github.com/glpi-project/glpi-agent/blob/1.17/Changes
About packaging, here is what you should retain:
As usual, we invite you to update your agents to take advantage of these improvements.
You can download it from the official project page on GitHub: https://github.com/glpi-project/glpi-agent/releases/tag/1.13
This version brings a few fixes and improvements, including:
proxy to none to not use this proxy system configurationrequired-category setting to help use GLPI business rules when full inventory reporting is enabled. This forces inclusion of a category in partial inventory even if content has not changed.itemtype and esx-itemtype configurations if the server supports generic asset types like GLPI 11+ willRegarding packages, here's what to note:
user and password were for proxy authentication; we corrected this and clarified the corresponding installer window to show they are for basic HTTP authentication, a common method supported by GLPI-Network CloudYou can check the full list of changes in the official Changelog available here: https://github.com/glpi-project/glpi-agent/blob/1.13/Changes
As usual, we recommend updating your agents to take advantage of these improvements.
GLPI Agent 1.11 has been released.
You can download it on the GLPI Agent github project: https://github.com/glpi-project/glpi-agent/releases/tag/1.11
This version comes with few fixes and enhancements and here are the essential ones:
ssl-keystore option is available to disable or optimize Keystore certificates export.full-inventory-postpone.glpi_protocol option is set and no GLPI server is configured.About packaging, here is what you should retain:
AGENTMONITOR_NEWTICKET_SCREENSHOT as parameter to configure if a screenshot should be done before opening a new ticket.You can check changes details in the official online Changelog available here: https://github.com/glpi-project/glpi-agent/blob/1.11/Changes
We invite you to update your agents to take advantage of these improvements.
GLPI Agent 1.8 has been released.
You can download it on the GLPI Agent github project: https://github.com/glpi-project/glpi-agent/releases/tag/1.8
This version comes with a lot of changes and enhancements for the MSI packaging for Windows:
AGENTMONITOR_NEWTICKET_URLto configure the GLPI url for the open new ticket option.On the inventory side, a new feature will help you reduce the carbon footprint of your GLPI system: the full inventory postpone
full-inventory-postpone is a new configuration parameter which value is 14 by default. Setting it to 0, it is still possible to disable this feature coming back to previous normal running.glpi-agent script is also supporting --full option to temporary disable the feature.The MacOSX package only activates Inventory task by default now. You’ll have to explicitly activate any other tasks you need in a dedicated configuration file.
A lot of fixes and enhancement have also been included. Inventory, RemoteInventory, NetworkDiscovery et NetInventory tasks, but also Proxy and ToolBox plugins, are concerned. We encourage you to check all changes details in the official online Changelog available here: https://github.com/glpi-project/glpi-agent/blob/1.8/Changes
We encourage you to update your agents as soon as possible to take advantage of these improvements.
GLPI Agent 1.6.1 has been released.
You're encouraged to upgrade your GLPI agents or migrate if you're still using FusionInventory agents.
You can download it on the GLPI Agent github project: https://github.com/glpi-project/glpi-agent/releases/tag/1.6.1
The 1.6.1 version specifically fixes SSL connections problems introduced with 1.6 version update for windows and MacOSX agents.
Here is a summary of the most important changes of the 1.6 version:
log, var and etc folders has been fixed during uninstallation,As always, you can check the more detailed changelog at: https://github.com/glpi-project/glpi-agent/blob/1.6.1/Changes
Very soon the new major version of GLPI will be released with many new features, including a major overhaul of the interface.
Here is the quick look at automatic inventory features.
Fork of the Perl FusionInventory agent, this new agent is enhanced with several new features:
The GLPI agent can advantageously replace the FusionInventory agent because it remains fully compatible with the automatic inventory of the FusionInventory.
This new agent is already available, you can download and test it:
https://github.com/glpi-project/glpi-agent/releases
Nightly built packages are also available to get the latest developments: https://nightly.glpi-project.org/glpi-agent/
We provide the community with a documentation detailing the installation, use and configuration of this new agent. Note that there may still be a few aspects to be completed (the missing points will be filled in the coming weeks).
We also provide a perl script (see dedicated documentation) allowing you to install agent on Linux with a suitable package (rpm, deb, snap.)
Some screenshots to illustrate the ToolBox interface of new agent:
Some time ago the android agent has returned to the google play store. It supports versions 4 to 10.
For GLPI Network subscription customers, we have also added the possibility of configuring the server URL via mechanism called "Deeplink". Using QRCode agent automatically retrieves the configuration to connect to the server.
We are happy to announce we add support for automatic inventory directly to GLPI core.
New API REST (front/inventory.php) will be available to receive inventory files in historical format (OCS, fusioninventory) or via new format json.
We therefore keep compatibility with the previous agents (just it will be necessary to reconfigure the server URLs of the historical agents, or to set up a web proxy to perform a redirection to the new URL).
You can now manage with this new format most of GLPI's inventory objects, like telephones, applications, racks, etc.
This is enabled by a new rewrite of the underlying code. The interface is inspired by the community projects mentioned above, the code is new.
This code now allows you to receive partial inventories.
This consists of sending only part of the particular information related to the object, and indicating to the server - via a flag - to update only the data concerned.
Transfer, processing, insertion and update of data is done in a significantly faster way.
As seen above, GLPI now integrates the basic building block concerning automatic inventory.
However we have not yet taken over the advanced tasks allowed by previous projects.
Here are a few points about these tasks:
For these last two features several scenarios are possible:
In our roadmap it is planned to work on rewriting of these parts, but as long as version 10 being almost done (and this substantial work), we are postponing their availability for a few months.
The main news of this version is a complete redesign of the interface and the user experience. The latest overhaul, dating back to version 0.90 was released in October 2015.
Based on tabler, Bootstrap 5 and Twig, this novelty brings many visual changes.
The integration of these standard libraries will make it easier to develop new interfaces and above all provides a “responsive” display mode that is easier to maintain.
Here are some screenshots to give you a preview:
We also have made a certain number of improvements of the user experience within GLPI. Without getting into the details, here is the list of key points.
We hope you enjoy these changes.
As you may have noticed with the previous captures, GLPI has now a default layout with a vertical menu on its left. If you still prefer a layout similar to previous versions (named Horizontal) you can find it in the the user preferences.
Now t is possible to collapse the menu in order to have a compact display.
Present since version 9.2 of GLPI, the “Go to” button allows a quick search of a menu. Now you can find it on top of the menu, its keyboard shortcut recalled.
The saved searches panel now is integrated to the GLPI design.
It can either be floating, as before, or pinned to pages, to be kept permanently (even during navigation).
Pinning is done on a page-by-page basis, a pinned panel for tickets will not be pinned for computers.
Note that it is now organized to present in the first tab, the saved searches corresponding to the type of object concerned: the list of tickets will display in priority order the saved searches concerning the tickets.
The second tab lists all the other searches concerning the other types of objects.
As the panel is now contextual to the type of object presented, access is via an icon named “Lists” present next to the breadcrumb.
It is now possible to disable the display of the search engine.
Very practical because you can navigate in GLPI only thanks to your saved searches!
The display of the search engine is now launched in “AJAX” mode. A loading icon is displayed when starting a search, changing pages or sorting. Once the new results are available, only the content is replaced and not the entire page.
With a “ctrl” you can add several sorts on the columns of your search results.
Example: sort by “Name” and by “Date modified”.
The main element form now displays a panel on the right recalling the images associated with the corresponding model.
These images were previously used only in the display part of data center racks, they are now available everywhere.
With the arrival of new libraries tabler and Bootstrap 5, it is now much easier to create and modify the palettes defining the colors of GLPI.
Here is sample palette file corresponding to the screenshot below:
2 new dark palettes are also available:
The display and use of ITIL objects (Tickets, Issues and Changes) were redesigned.
Many changes were made:
The user who asked the question will be added as an observer of the object and a notification will be sent to him personally.
A new notification (and a corresponding event) named “New user mentioned” is added to your GLPI instance.
You can now define the timeline for a follow-up.
In addition, an automatic reminder after a certain period and with recurrence can be defined in order to warn participants that a response from them is expected. It is also possible to plan an automatic resolution after a certain number of reminders.
This display mode previously only available in projects, is now accessible from the Helpdesk via dedicated icon next to the breadcrumb.
You can add columns corresponding to the statuses.
Attention, even if the “Closed” status is available, we will not display the relative data in order to prevent the execution time from taking too long. The column remains available to accommodate the object move.
On this new page and also on the page of projects it is now possible to add criteria to filter the cards.
Follow-ups, tasks and solutions templates now support the addition of variables. These are replaced by their values at the time of their use. For example, it is possible to use {{ticket.id}} in a template and when choosing the template in a follow-up, the variable is replaced by the ID of the current ticket.
Integrated documentation is available to know all the variables available.
We now support automatic inventory directly in GLPI.
A new REST API will be available (directly in the GLPI index) to receive inventory files in historical xml format or via the new json format.
We therefore maintain compatibility with previous agents (it will just be necessary to reconfigure the server URLs of historical agents, or to set up a web proxy to redirect to the new URL).
We now manage with this new format most of the GLPI inventory objects, such as telephones, applications, racks, etc.
This code now allows you to receive partial inventories.
This consists of sending only part of the information related to the object concerned, and indicating to the server, with a key, to update only the data concerned.
Transporting, processing, inserting and updating data is done significantly faster.
This new agent (bifurcating since version 2.6 of the fusioninventory project) has been enhanced with several new features:
You can get the archives for your systems at the following address: https://github.com/glpi-project/glpi-agent/releases
Nightly build packages are also available to get the latest developments: https://nightly.glpi-project.org/glpi-agent/
We also provide a perl script (see documentation dedicated) allowing to install the agent on your linux with a suitable package (rpm, deb, snap.)
The documentation is available and details of the installation, use and configuration of this new agent.
Some screenshots to illustrate the ToolBox interface of this new agent:
The android agent is back on the google play store. It supports versions 4 to 10 of the OS.
For customers under subscriptions we have also added the possibility of configuring the server URL via a “Deeplink”. Via QRCode, the agent automatically retrieves the configuration to connect to the server.
As seen above, GLPI now includes the basic brick for automatic inventory.
However we have not yet resumed the advanced tasks allowed by the previous projects.
Here are the points about these tasks:
For these last two features, two temporary scenarios are possible:
This package is transitional and the integration of these modules is planned in our roadmap, the complete package is expected in a few months.
We have started the work on developing the remote deployment module. In addition to its integration into the core of GLPI, a temporary community plugin will be made available to use the functionality before the next major version of GLPI.
We will continue SNMP and ESX functionality in the same way during the year: support available in the next major release, and community plugin for beta testing of the functionality.
Contribution of Fian services and Smartwood, the old Gantt view has been refreshed by integrating the library DHTMX Gantt.
This change also brings an interactivity of the view (possibility to create/modify/move tasks).
Material (items) reservations now use the same library (fullcalendar) as GLPI planning.
From the preferences, it is now possible to choose for the rich text fields (html), the layout of the integrated toolbar:
In the settings it is now possible to choose for the rich text fields (html), the layout of the integrated toolbar:
The “embedded” layout displays a contextual embedded toolbar to the right of your cursor.
Selecting text will bring up other options such as bold or italic.
Other less common options (source code, full screen, etc) will be available via right click.
The plugin databases previously carried by Infotel is now directly integrated (Management menu) thanks to their help.
A migration script to recover data from the old plugin is available from the command line console.
These two objects now can be linked, tabs have been added to them.
A new cable object is now added to the Assets menu.
It is used to represent the physical link between the ports of two inventory objects.
New prerequisites:
If you are ready to try it, download the final version here:
