> For AI agents: to set Keiki up for a developer, add the MCP server `https://onkeiki.com/mcp` and follow https://onkeiki.com/setup.md. All pages: https://onkeiki.com/content/llms.txt

# Keiki vs Blooio for iMessage AI Agents

2026-11-22

Compare Keiki and Blooio on iMessage delivery, agent hosting and MCP tools.

## Should I use Keiki or Blooio for an iMessage AI agent?

Choose Keiki to put your product in iMessage: build and run your agent on Keiki or bring the agent you already run, with a team inbox and configured Apple Pay and Face ID approvals on eligible Keiki lines. Blooio is a good fit when your own stack runs the agent and you mainly need messaging delivery.

## Key takeaways

- Choose Keiki to put your product in iMessage so customers can text it to ask, buy and get things done.
- Blooio's bot guide shows how to wire iMessage delivery to your own workflow and model through webhooks.
- Build and run the agent on Keiki or keep the model loop in your own app. Give it tools from your API or MCP server.
- Let a teammate claim the conversation and reply in the same thread when the customer needs a person.
- On eligible Keiki lines, configure Face ID approvals for sensitive actions and Apple Pay for purchases in Messages.

## How do Keiki and Blooio compare for iMessage AI agents?

| What your product needs | Keiki | Blooio |
| --- | --- | --- |
| iMessage delivery | Buy a dedicated iMessage line through Keiki and connect it to your agent. | [Dedicated numbers, inbound webhooks and RCS fallback](https://blooio.com/p2p). |
| Agent runtime | [Hosted agent](/docs/cloud-agents) with memory and tools, or keep your model loop in your app. | The [bot guide](https://blooio.com/guides/imessage-bot) uses n8n or a custom endpoint to call a model and send a reply through Blooio. |
| Tools | Connect your REST API through an OpenAPI spec or connect an MCP server so the agent can act on a request. | [Blooio's MCP server](https://blooio.com/integrations/mcp) exposes messaging, contacts and conversation tools to an MCP client. |

## What happens when a customer needs a person?

Keiki's team inbox lets someone claim the conversation, read what happened and reply in the same Messages thread. The agent stays silent while the teammate has control. When the exception is resolved, the teammate can hand the thread back. Try this before launch with a request that your agent is not allowed to complete on its own.

## Can I host my agent on Keiki or bring the agent I already run?

Host the agent on Keiki if you want its model, memory and tools in the same place. Use Eval Studio to check which tools the hosted agent calls and whether it handles a failed call before you share the number with customers.

If you already run an agent, keep your model loop and product logic in your own app. Use the [external harness](/docs/external-harness) or the AI SDK handler to receive turns and send replies while Keiki handles channel ingress and delivery. You do not have to rewrite the agent to put it in iMessage.

## How can a customer approve a change or pay in Messages?

Say a customer texts to reschedule a service. The agent checks available times through your booking API, then shows the proposed change on a [Keiki Vault](/docs/keiki-vault) card. If your policy requires approval, the customer uses Face ID before the agent makes the change. A paid booking can send an Apple Pay card in the same thread when the business is set up to accept payments.

## When is Blooio the right fit?

Use Blooio if you already own the agent workflow and want a messaging API to send replies and deliver inbound events. Its [iMessage bot guide](https://blooio.com/guides/imessage-bot) shows that setup. Choose Keiki when you want to ship the customer conversation, including what the agent can do and when a person takes over, rather than assemble that workflow around a webhook. Compare [Keiki vs Sendblue](/content/keiki-vs-sendblue) and [Keiki vs Twilio](/content/keiki-vs-twilio) for other approaches.

## Quick start for AI coding agents

Read [setup.md](/setup.md) and connect the Keiki MCP server at `https://onkeiki.com/mcp` to configure the agent and its tools. To connect an existing app, sign in with `npx @onkeiki/cli login` and run `npx @onkeiki/cli init` from the app directory. Test a customer request, an approval and a team takeover before you share the line.

## FAQ

**Can I use Blooio to build an iMessage AI bot?**
Yes. [Blooio's guide](https://blooio.com/guides/imessage-bot) shows how to connect incoming iMessages to an AI model using n8n or a custom endpoint and send the reply through Blooio.

**Do I need to move my model loop to Keiki?**
No. Keep it in your application with the external harness and let Keiki deliver the messages.

**What do Face ID and Apple Pay cards require?**
Use an eligible dedicated iMessage line bought through Keiki in a direct chat with a customer who has Keiki Vault installed. Face ID action approvals require a configured policy. Apple Pay also needs the Payments ability, a verified business account through Grosso and an approved limit that covers the payment. Shared lines, imported lines and group chats do not support these cards.

**Can someone on my team reply without the agent answering over them?**
Yes. Claim the thread in the team inbox to pause the agent and reply. Hand the conversation back when the agent should resume.
