
Intelligent Search finds messages by meaning across your entire workspace, with filters that work in air-gapped environments, indexing that scales with message volume, and admin controls that keep results precise and trustworthy. Rocket.Chat now fully conforms to WCAG 2.1 Level AA across web, iOS, and Android, independently verified by two external auditors and satisfying Section 508, the European Accessibility Act, BITV 2.0, and EN 301 549.
The organizations we build for operate in environments where two things are non-negotiable: the ability to act on the knowledge they already have, and the ability to prove they meet the standards required of them.
The latest version of Rocket.Chat advances both.
- Intelligent Search makes your workspace's institutional knowledge retrievable, not just stored.
- And Rocket.Chat's accessibility conformance is now independently certified, publicly documented, and built to hold up in any procurement review.
Here is what is new in v8.5.
Intelligent Search: Surface the right information, exactly when it matters
Intelligent Search finds messages by meaning, not by exact keyword match. You describe what you are looking for, and it returns the most relevant results from across your workspace, ranked by how closely they match the intent behind your query, not just when they were sent.
The latest version of Intelligent Search improves how users search and how administrators manage search across their workspace.
For users
Find exactly what you need, without the noise
Searching across a large workspace can feel like looking for a specific conversation in a room where everyone is talking at once. Filters change that.
When supported by the workspace configuration, you can filter any search by username, channels or time windows.
The result is fewer irrelevant matches and faster access to what you actually need:
- @username: Find everything a specific person said on a topic, across every channel you have access to.
- A procurement lead searching @sarah vendor evaluation gets every relevant message Sarah sent, whether in a project channel, a direct message, or an incident thread, without manually checking each one.
- #room: Scope results to a specific channel.
- An engineer joining mid-incident who searches authentication errors #incident-response gets full context immediately, rather than scrolling through hundreds of messages to piece it together.
- my messages / from me: Retrieve things you said.
- A project lead searching my messages delivery timeline finds the exact commitment they made weeks ago, right before a client follow-up.
- today, yesterday, this week: Look for messages within a specific time window.
- A compliance officer can surface everything discussed on a specific regulatory topic that day, across all accessible channels, in a single query rather than checking each channel individually.
For organizations under GDPR, FedRAMP, or operating in classified environments, every filter runs entirely within your infrastructure. No query is routed to an external model. No data leaves your environment, so search works without compromising how you handle sensitive information.
Spend less time searching and more time acting on what you find
When a search returns no results, the app now shows the exact query it used. You see what was searched, understand where it missed, and know precisely how to refine it. You are never left guessing why something did not come back.
First-time users receive clear guidance on how semantic search works and how to use filters, so everyone on the team gets value from their first search, not their fifth.
Direct message results show you who the conversation was with, immediately
Results from direct message conversations now display participant names as DM: user, user, rather than a system-generated ID. You know instantly whether you have found the right exchange, without clicking into it to find out.
For administrators
You decide what gets searched
Administrators now have clear visibility into which rooms are indexed live, which are in historical backfill, and which are out of scope. This matters because a searchable knowledge base is only valuable if it contains the right information. You can index selectively and know exactly what Intelligent Search will and will not surface, rather than indexing everything and leaving users to filter through noise.
Results your users can trust
- Bot-generated messages are excluded from indexing automatically. When users search, they get results from real conversations, not automated system responses that add noise without adding value.
- Room references in queries are validated against what a user actually has access to, so search results always respect your existing permission structure. Users only ever see what they are supposed to see.
Search results tuned to your organization
Every result Intelligent Search returns has a relevance score: a measure of how closely it matches the meaning of the query. The minimum relevance threshold is the cutoff below which results do not appear. Administrators can now configure this threshold per workspace.
- Set it high and only the strongest matches surface, fewer results, but each one closely aligned with what was searched. A support team that needs precise answers to specific technical questions benefits from a tighter cutoff.
- Set it low and the net widens, more results appear, including messages that are loosely related, which is useful when the goal is discovery rather than pinpoint retrieval. A team building a knowledge base or exploring historical decisions benefits from broader coverage.
Search performance that holds as your workspace scales
For Intelligent Search to work, messages need to be indexed: read, processed, and organized in a way that makes them searchable by meaning.
Previously, indexing ran inside the same application handling live search, meaning the two competed for resources. In this latest version, indexing operates independently. A workspace processing thousands of messages in the background delivers the same search experience to users as one that is idle. Performance does not degrade as your workspace grows.
Accessibility: Full WCAG 2.1 Level AA conformance, across web, iOS, and Android
With 8.5, Rocket.Chat achieves full conformance with WCAG 2.1 Level AA across the web application, iOS, and Android, on the timeline publicly committed to in our Accessibility Statement. That conformance is independently verified through two external audits.
The organizations we serve operate in environments where accessibility is not optional. It is a legal requirement, a procurement condition, and a commitment to the people using these tools every day.
We take that seriously, which is why this milestone is backed by independent verification, current documentation across web, iOS, and Android, and a program that continues well past this release.
Independently verified by two external auditors, not self-declared
We wanted the work to be verifiable, not just visible. CCNET in Germany, a government-authorized auditing body, certified Rocket.Chat against BITV 2.0. A separate independent engagement was conducted by auditors in Norway. Every barrier identified across both audits is now closed.
The standards this conformance satisfies:
- WCAG 2.1 Level AA: the international baseline for digital accessibility
- BITV 2.0: Germany's binding national implementation of the EU Web Accessibility Directive, certified by a government-authorized auditor
- EN 301 549: the EU harmonized standard for ICT accessibility
- EAA (Directive 2019/882): the European Accessibility Act, enforceable across all EU member states since 28 June 2025
- Section 508: the US federal requirement for accessible information and communication technology
The EAA is in effect
The European Accessibility Act became enforceable on 28 June 2025. If your organization is working through procurement reviews that require EAA compliance, Rocket.Chat's BITV 2.0 certification gives you an independently verified, government-authorized answer to that requirement today.
An ongoing program, not a one-time milestone
We see accessibility as a standard of quality, not a destination. Rocket.Chat maintains a dedicated accessibility team, runs recurring audits across web, iOS, and Android, and keeps VPATs for web, iOS, and Android current.
When new barriers are identified, they are addressed and reflected in updated documentation. Organizations that need to demonstrate continuous compliance, not just a one-time conformance claim, can rely on the program to support that.
Documentation for procurement
Other enhancements
Pick up exactly where you left off with Drafts
- When you start typing a message and switch rooms before sending, that draft is now saved and visible in a dedicated Drafts group at the top of your sidebar.
- Select the room and your message is right there, ready to continue.
- Once sent or cleared, it disappears from the group.
- Enable it under Account > Preferences > Feature Preview.
- Audio and video recording also stays available while you have text in the composer, so you can switch to a voice or video message without losing what you typed.
Please note: Each room appears in only one group at a time, so your sidebar stays organized: Unread takes priority over Drafts, and Drafts over Favorites.
Tighter control over who manages what, and what stays protected
- The ABAC administration page is organized into tabs, now each with its own permission set.
- An administrator responsible for room policies does not see or touch schema management controls, and vice versa. This reduces the risk of unintended changes across a sensitive access control system.
- When an ABAC-managed room is converted to a public channel, its access attributes are removed automatically, so restricted metadata does not carry over to a room that is now open to everyone.
Bug fixes
This release includes a number of fixes to improve overall stability, performance, and user experience. For a complete list of resolved issues in Rocket.Chat v8.5, check out the official release notes.
How to update your workspace to the latest Rocket.Chat version?
SaaS workspaces
The release happens automatically on our cloud, so no further action is required from your side. However, remember that updating instances might take a few weeks, so don't hesitate to contact our support team if you need to update your version sooner.
Self-managed workspaces
Depending on the installation mode, you might need to update your server manually. Check out the documentation for instructions.
Frequently asked questions about <anything>
- Digital sovereignty
- Federation capabilities
- Scalable and white-labeled
- Highly scalable and secure
- Full patient conversation history
- HIPAA-ready
for mission-critical operations
- On-premise and air-gapped ready
- Full control over sensitive data
- Secure cross-agency collaboration
- Open source code
- Highly secure and scalable
- Unmatched flexibility
- End-to-end encryption
- Cloud or on-prem deployment
- Supports compliance with HIPAA, GDPR, FINRA, and more
- Supports compliance with HIPAA, GDPR, FINRA, and more
- Highly secure and flexible
- On-prem or cloud deployment



