# AI agents are capacity, not seats. Canonical HTML: https://www.praxis-agents.ai/articles/capacity-not-seats Stop counting heads. Start buying capability. Most software charges per person. AI agents don't work like that. Praxis Agents are shared capacity - one system your whole team builds on. No per-head pricing. No licence negotiations every time you hire. ## You're buying capability, not logins Seat-based pricing makes sense for individual tools. But agents work across your team, so the economics are different. ### Seat-based software #### You're paying for logins. That makes sense for tools where each person gets their own workspace. CRM, design tools, project management - the value is tied to the individual using it. ### Praxis Agents #### You're paying for what gets done. The system sits across your team. It handles work, remembers decisions, and keeps things moving. The value isn't in who logs in - it's in what the business gets out of it. ### What changes for people #### Your team does higher-value work. When the repeatable stuff is handled reliably, people spend less time chasing and more time on the things that actually need a human - decisions, relationships, exceptions. ## Why the seat analogy doesn't fit Seat pricing works when each person gets their own tool. Agents are different - they sit underneath the team, handling work that crosses people and functions. The biggest wins don't come from one person using one interface. They come from how work gets prepared, handed off, checked, and followed through - across the whole team. ### What you're paying for - Seat model: A login for each person - Capacity model: A shared layer that the whole team benefits from ### What grows the bill - Seat model: Headcount - more people, more licences - Capacity model: Usage - more work through the system, more infrastructure ### Where the value shows up - Seat model: One person using one tool - Capacity model: Better handoffs, fewer dropped balls, more consistent output across the business ### When someone leaves - Seat model: The tool stays, but a lot of the working knowledge walks out the door - Capacity model: The instructions, workflows, and decisions stay with the company ## Knowledge that stays when people move on The real advantage isn't raw automation. It's that instructions, workflows, and past decisions don't have to be rebuilt from scratch every time someone leaves or a project changes hands. This matters commercially because the value compounds. You're not just getting output today - you're keeping the operational knowledge that normally ends up scattered across inboxes, documents, and people's heads. New starter? They inherit a working system. Someone leaves? The structure stays. That's capacity, not a seat. ### What stays with the company #### Retained working context - Instructions and business rules - Workflow steps and approval chains - Past decisions and the reasoning behind them - Connections to your existing systems - Drafts, outputs, and completed work - Operational knowledge that compounds over time ## The value shows up everywhere, not just on one screen If you're measuring this by how much time one person saves, you're looking at the wrong thing. ### Cleaner handoffs Work doesn't need to be re-explained every time it moves between teams. The context travels with it. ### Faster onboarding New starters inherit a working system, not a folder of documents and tribal knowledge that takes months to absorb. ### More consistent execution The same standards, checks, and processes run every time - not just when the right person happens to remember. ### No seat creep More people benefit from the same system without every new hire triggering another licence negotiation. ## Not fewer people. Better use of people. Agents pick up the repeatable stuff so your team has more room for the work that actually needs them. ### Humans focus on - Judgement calls and trade-offs - Client relationships - Commercial decisions - Creative direction - Setting priorities - Handling sensitive exceptions ### Agents support with - Pulling together information - First-draft writing and formatting - Following up and chasing - Keeping systems in sync - Running routine checks - Remembering what happened last time ## Next step ### Let's talk through the numbers If you want to understand how Praxis Agents would work commercially for your team, we can walk through it together. [Get in touch](/contact.md) ## Related ### Why this isn't SaaS The pricing makes more sense once you see why Praxis Agents aren't a shared platform in the first place. [Read the article](/articles/why-not-saas.md) ## FAQ ### Is this just SaaS with different branding? No. SaaS charges per user because the value is tied to individual access. Praxis Agents are shared infrastructure - more like hiring capacity than buying software licences. The whole team benefits from the same system. ### Why does it matter that the system retains knowledge? Because the workflows, rules, and decisions your team builds into the system stay there. When someone leaves or a project changes hands, that knowledge doesn't disappear. It compounds. ### How does pricing actually work? A platform fee plus the underlying cloud and AI costs for what you actually use. Your bill grows with usage, not headcount - so adding another person to the team doesn't automatically mean paying more. ### Will this replace people on my team? That's not the goal. Agents handle the repeatable operational work so your team can focus on the things that genuinely need a human - judgement, relationships, creative thinking, and the decisions that move the business forward.