> ## Documentation Index
> Fetch the complete documentation index at: https://docs.eat-now.io/llms.txt
> Use this file to discover all available pages before exploring further.

# Plans

> Five plans — from free trial to high-volume — built around the number of WhatsApp conversations the AI agent handles for you each month.

The AI agent is a paid add-on on top of your EatNow subscription. You pick a tier that matches your reservation volume, and what counts each month is **conversations** — not messages, not turns, not credits.

<Note>
  **One conversation** = one WhatsApp thread in which the AI agent sends at least one reply, within a 24-hour window. A customer who books, then comes back the next day to modify, counts as **two** conversations.
</Note>

## Plans

<CardGroup cols={2}>
  <Card title="Free" icon="gift">
    **20 conversations / month**

    <br />

    To try out the agent on real customer messages. Goes silent (hands off to your team) once the quota is used.
  </Card>

  <Card title="Starter" icon="seedling">
    **200 conversations / month**

    <br />

    A small restaurant (≤ 100 bookings / month) that wants the agent to handle the easy stuff.
  </Card>

  <Card title="Growth" icon="chart-line">
    **750 conversations / month**

    <br />

    A growing restaurant (200–400 bookings / month) that needs more headroom than Starter.
  </Card>

  <Card title="Pro" icon="rocket" iconType="solid">
    **1,500 conversations / month**

    <br />

    The standard plan for an active restaurant (500–700 bookings / month).
  </Card>

  <Card title="Enterprise" icon="building">
    **5,000 conversations / month**

    <br />

    Large rooms, groups, or restaurants with a heavy inbound.
  </Card>
</CardGroup>

<Tip>
  Talk to your EatNow contact for the up-to-date pricing of each plan and any custom volume you may need.
</Tip>

## What a conversation looks like

A typical booking conversation runs about 4–6 messages and lasts a few minutes. From the agent's side, that's **one conversation** counted toward your quota — regardless of how many times back and forth.

<CardGroup cols={2}>
  <Card title="Booking inquiry" icon="message-circle">
    "Hi, do you have a table for 4 on Saturday at 8pm?" → agent checks availability, asks for an email, confirms. **1 conversation.**
  </Card>

  <Card title="Modification" icon="edit">
    "Hi, it's Lisa — can we move our Saturday booking to 9pm?" the next day. **1 conversation** (separate from the original booking).
  </Card>

  <Card title="Quick question" icon="circle-help">
    "Do you have a parking?" — agent answers from your knowledge base. **1 conversation.**
  </Card>

  <Card title="Late arrival" icon="clock">
    "Running 15 min late!" — agent adds a note to your reservation. **1 conversation.**
  </Card>
</CardGroup>

The agent doesn't take breaks, doesn't sleep, and replies in seconds — including at midnight or during a packed Saturday service when your team has zero capacity to answer the phone.

## What if I hit the quota?

You decide ahead of time, in your settings:

<Tabs>
  <Tab title="Hard stop (default)">
    Once the quota is used up, the agent stops replying and **escalates new conversations to your team** until the next monthly refill. You'll never be billed for more than your plan.
  </Tab>

  <Tab title="Continue on overage">
    The agent keeps answering past the quota; every extra conversation is added to your next invoice. Switch it on if you'd rather take the higher bill than have a quiet thread.
  </Tab>
</Tabs>

Quotas refill on the **1st of each month**. Unused conversations don't roll over.

## FAQ

<AccordionGroup>
  <Accordion title="Do customer-sent messages count?">
    No. The counter only ticks when the agent **sends a reply**. A customer who messages but is ignored (e.g. outside your enabled hours, blocked by a safety rule, or after a hard stop) does not consume a conversation.
  </Accordion>

  <Accordion title="What happens if the agent escalates mid-conversation?">
    Still **one conversation**. The first agent reply in a 24h window is what locks the counter — escalation or hand-off afterwards is included.
  </Accordion>

  <Accordion title="Can I change plans mid-month?">
    Yes. The new quota applies right away from your dashboard's plan-change moment. You don't lose what you've already consumed, but your new balance reflects the new plan's quota.
  </Accordion>

  <Accordion title="What about the WhatsApp Business API fees from Meta?">
    Included. The conversation count covers both the agent's compute and the Meta-side cost — there's no separate WhatsApp bill on your side for the agent's replies.
  </Accordion>

  <Accordion title="Can I disable the agent without losing my plan?">
    Yes. You can pause the agent from your settings any time. Your quota doesn't tick while it's paused, and the plan stays active so you can resume in one click.
  </Accordion>
</AccordionGroup>

<Tip>
  Not sure which plan fits? Start on **Free** to test the agent on your real conversations. The Usage tab in your dashboard shows you exactly how many conversations you'd consume on a normal month — pick a plan after a week of data, not before.
</Tip>