Xero announced XeroForce on 14th May 2026, its own way to build AI agents on top of your accounting data. It's a genuine step forward, and it puts a real question in front of finance teams: should your AI agents live inside Xero, or should your own AI reach into Xero from the outside?
One practical note before the comparison. XeroForce launched as an invite-only alpha, with general release planned for later in 2026. So for most businesses this isn't yet a choice you can act on, it's a direction worth understanding. This is an honest look at what each approach is, where each fits and how to decide when both are available to you.
What Is XeroForce?
XeroForce is Xero's no-code AI agent builder. You describe what you want in plain language and it creates an agent that runs on top of your Xero data and the third-party apps you connect, with accounting and financial logic already built in. It's powered by Xero OS, the AI-native platform layer underneath Xero's products, and every action is logged and traceable for compliance. It also handles long-running background processes, so an agent can wait on an event like an email reply or a filing date, and gives accountants bulk actions across the clients they manage.
The strategic idea is clear from Xero's own framing: Xero wants to be the orchestration hub, the core financial operating system your agents run inside. XeroForce is how you build those agents, and JAX ("Just Ask Xero"), Xero's financial superagent, is how it orchestrates them. For a business that runs its finances entirely in Xero, that's a tidy, well-integrated story.
What Is a Xero MCP?
A Xero MCP takes the opposite direction. MCP (Model Context Protocol) is an open standard that lets an AI tool connect to an external system. A Xero MCP server exposes your Xero data to whichever AI you already use, so your agent isn't something you build inside Xero, it's the AI your team already works in reaching into Xero to do the work.
Xero publishes an official MCP server for single-entity use. ELab AI's MCP integration for Xero is built for businesses running multiple Xero organisations: it connects every organisation you have authorised from one session, runs on hosted infrastructure, and adds stronger invoice and credit note search, Xero Projects support and governed access. The AI doing the work can be Claude, ChatGPT, Cursor or any MCP-compatible tool.
The Core Difference: Whose AI Runs Your Finance Work?
Strip away the detail and it comes down to one question. With XeroForce, the agents live inside Xero's ecosystem and run on Xero's platform. With a Xero MCP, your own AI lives wherever you already work and reaches into Xero to get things done.
That single distinction drives everything else. Agents built in XeroForce are as portable as Xero itself, which for a Xero-first business is fine. Work done through an MCP stays with your AI tool and your own workflows, so it travels with you if your stack changes or if different parts of your team use different AI.
How the Two Approaches Compare
| Dimension | XeroForce | ELab AI MCP Integration |
|---|---|---|
| Availability | Invite-only alpha, general release planned for later in 2026 | Available now |
| Where agents run | Inside Xero, on Xero OS | In your own AI tool (Claude, ChatGPT, Cursor) |
| Cross-entity work | Bulk actions across an accountant's clients, but not positioned for one group's own entities | Every authorised organisation queryable in one session |
| Which AI does the work | Xero's | Whichever you already use |
| Platform lock-in | Tied to the Xero ecosystem | None: MCP is an open standard |
| Search and filtering | Whatever your agent is built to do | Invoice and credit note filters by reference, status and date range |
| Audit trail | Yes | Yes |
| Best suited to | Single-entity businesses run entirely in Xero, and accountants managing many clients | Multi-entity finance teams using their own AI |
When XeroForce Is the Right Choice
XeroForce fits well if you run a single Xero organisation and do most of your finance work inside Xero already. If you don't use a separate AI tool and don't want to, building agents where your data lives is the simplest path. The no-code builder means you don't need technical help to get started, and the accounting logic being baked in removes a lot of setup.
It also looks strong for accounting practices. Bulk actions across the clients you manage, plus agents that can sit waiting on a filing date or an email reply, map neatly onto how a practice actually works.
For a small business or a practice working inside Xero, XeroForce is a sensible option from the vendor that already holds your data. The caveat is timing: you'll need an invite until general release.
When a Xero MCP Is the Right Choice
A Xero MCP fits when one or more of these is true:
- You run multiple entities. Property groups, investment firms and multi-brand operators hit the single-organisation wall quickly. An MCP that connects every organisation in one session lets you search and work across them without switching.
- Your team already works in Claude, ChatGPT or Cursor. If your finance analysis, drafting and research already happen in an AI tool, reaching into Xero from there is less friction than moving into a separate builder.
- You don't want your workflows locked to one platform. Building your finance automation on an open standard means it isn't stranded if your tools or platforms change.
What About JAX?
JAX, Xero's "Just Ask Xero" superagent, sits in the same ecosystem logic as XeroForce. It's designed to orchestrate agents and answer questions inside Xero. It's a strong feature for Xero-first businesses, and it reinforces the same trade-off: it's excellent within the Xero ecosystem, and it keeps the intelligence there rather than in the AI tools your team might use across the rest of the business.
The Bottom Line
Neither approach is simply better than the other. XeroForce suits single-entity businesses and accounting practices that want everything inside Xero, once general release arrives. A Xero MCP suits finance teams running multiple entities who want to use their own AI and keep their workflows portable, and it's available today.
If that second description sounds like your business, see how ELab AI's MCP integration for Xero works or read our guide to connecting Xero to Claude. If your team works in ChatGPT or Microsoft 365 Copilot instead, those setups have their own pages.
Frequently Asked Questions
Is XeroForce the Same as a Xero MCP?
No. XeroForce is Xero's own no-code builder for creating AI agents that run inside Xero. A Xero MCP is a connection that lets an external AI tool, such as Claude or ChatGPT, work with your Xero data from outside the Xero ecosystem.
Does XeroForce Support Multiple Xero Organisations?
XeroForce includes bulk actions for accountants working across the clients they manage, but Xero hasn't positioned it around working across the entities of a single group. If you need to search and work across your own organisations in one session, a multi-organisation Xero MCP is built for that specific problem.
Is XeroForce Available Yet?
At launch in May 2026 it was an invite-only alpha, with Xero planning general release later in the year. Check with Xero for current availability in your region.
Can I Use XeroForce With Claude or ChatGPT?
XeroForce agents run inside Xero on Xero OS, not in external AI tools. To use Claude, ChatGPT or Cursor with your Xero data, you connect through a Xero MCP instead.
Is XeroForce Free?
Xero hasn't published pricing alongside the alpha. Since it's part of the Xero platform, cost and access are likely to depend on your plan and region, so check current details with Xero directly.
Which Approach Avoids Platform Lock-in?
A Xero MCP built on the open MCP standard keeps your finance workflows portable across AI tools. XeroForce keeps them inside the Xero ecosystem, which is convenient for Xero-first businesses but ties the workflows to that platform.