Rajiv Tirumalareddy, Thanks for the update, but I want to be candid: usage-based billing, while it sounds appealing on the surface, feels more like a workaround than an actual solution to what's being requested here.

The original request, and what this entire thread has been consistently reinforcing for over two years, is a specific per-user seat model. That's the correct, sustainable solution for any organisation that is cost-conscious. It provides predictability, clear ownership, and control: you know exactly who has AI access, and you pay only for those seats.

Usage-based billing shifts the problem rather than solving it. Without per-seat control, teams can't govern who uses AI, they can only react to how much was consumed after the fact. For organisations where only a handful of users genuinely need AI (which seems to be the vast majority of voices in this thread), that's not a long-term solution, it's just a different kind of unpredictability.

The per-user seat model isn't a niche request. It's industry-standard (GitHub Copilot, Notion AI, and others have proven this works). We'd strongly encourage ClickUp to reconsider putting it back on the near-term roadmap rather than building around it.

Hey everyone, thanks for all the continued discussion here. Wanted to drop a quick update.

The specific per-user seat model as originally requested is not on our near-term roadmap. But the good news is that over the next few months, we’re working on something that will shift a lot of AI features to usage-based billing, so you’re paying for what you actually use rather than how many seats you have. We think this will address the core pain point a lot of you have raised.

We’re not ready to share more details just yet, but we will soon. Really appreciate the patience and keep the feedback coming!

I believe using Brain AI on a credit-system basis workspace wide or a user-specific subscription pricing can be the best workaround. Hoping that this feature comes soon!

The reality of our organization (and it seems many others here) is that we have a 90/10 split:

10% Power Users: Managers/POs who desperately need AI for planning, writing, and organizing. We are ready to pay for these seats today.

90% Contributors: Users who just update status and comment. They do not need AI, and we cannot justify doubling our bill to give it to them.

By forcing us to buy for the 90% to enable the 10%, you are currently getting $0. A mixed model (Per-seat for active users + Credits for background tasks) allows us to start paying you immediately.

It’s actually quite simple. If we look at the real-world workflow of a software company, for example, not every operational team member needs AI, but middle management and upper leadership definitely do. Here are a couple of real-life cases:

Case 1: Product Owner and Product Creation

A user in the PO role could use AI to create ticket content much faster and more easily. This doesn't necessarily mean the person executing the ticket needs AI; the PO simply used it to refine user stories, epics, and milestones.

Case 2: Project Manager/Scrum Master and Cross-Functional Management

Analyzing data (which already exists in dashboards, Gantt views, tables, etc.) would be much easier if the user in this role uses ClickUp’s AI. They wouldn't have to export all the data to third-party tools or external AIs. Furthermore, they could get faster insights into deviations, blockers, and delays just by using the data ClickUp already has for every ticket, assignee, and project. In this case, it isn't necessary for more users to have AI access since they won't use it—the AI feeds on what already exists within ClickUp.

Case 3: Upper Management

A CEO or CTO being able to ask the AI, "What was the task completion rate for the product team during Q4 2025?" doesn't require the entire company to have AI. Once again, the AI already has access to all the data.

Ultimately, it’s easy when a product like ClickUp Brain is built based on user needs rather than just the bottom line.

Yes - we need this! Only some users require it

From my perspective, per-user AI billing still makes the most sense, even if it means being stricter with workspace-level features.

I’d be completely okay with Workspace-level AI features (like AI Assign / AI Prioritization) being restricted to spaces, folders, or teams where all users have an AI seat. That feels similar to how other tools gate advanced automations or permissions.

In practice, most organisations already have:

If a mixed model is needed, I’d strongly prefer:

Hi everyone, much like many others, in our organisation we'd only have likely 3-4 people using AI out of the 30+ we have in our Workspace. I think for this reason, per-seat billing for individual AI tools will work well combined with either an option choose a workspace wide set of AI features that are pre-built into the plan level as some organisations may not need all of the AI features OR the choice to use capped credits with clear and detailed reporting for credit packs incase they want all of the AI features but not to pay for a high cost plan. It would be great as someone also mentioned below to be warned before your credit pack runs out so that jobs don't fall through the cracks and get missed. I think the ability to mix and match needs is key. Thanks.

Rajiv Tirumalareddy admins should be able to decide who has access to prompt AI to perform tasks. The "owner" of the AI action is the person who set up or prompted the AI to perform the action.

For my company, as an example - Only team leads would need access to AI tools. The AI features are incredibly helpful for creating and assigning tasks to our developers. Our developers really only update their tasks, documentation and leave comments on tasks. They wouldn't need AI to perform their duties at all, but it would be good to have the ability to grant AI access to one or two developers who we think WOULD use the tool.

Paying for the entire workspace to have access to the AI tools when only a small percentage actually uses them is a waste of money.