How many AI replies your plan includes and how to estimate what you need

Omnifox AI explained in replies, not tokens: how many each plan includes, what happens when they run out, and an example with 1,000 conversations a month.

Oct 19, 2026

Watch this step in the video · AI tokens and top consumers

Omnifox AI agents are on every plan, starting with Lite. Each plan includes a monthly amount of AI. Internally it's measured in tokens (small pieces of text), but for planning it's easier to think in replies: each message the AI sends to a customer.

How many replies each plan includes

Plan AI replies per month (approx.)
Lite 600
Impulsa 1,200
Conecta 2,000
Crece 4,000
Escala 6,000
Exclusivo Custom

These are approximate figures: a short answer to a simple question uses less than a long answer that checks your knowledge base, reads the history and uses tools (booking an appointment, searching for a product). Other AI features, like summaries, audio transcription or translation, also draw from the same allowance.

The allowance resets at the start of each billing cycle.

What happens when it runs out

The AI doesn't stop as long as you have balance: it keeps replying and draws from your top-up AI wallet (the Omnifox balance), available on every plan. For that:

  1. Go to Settings → Billing, Balances tab, and click Add credits to top up. You can turn on auto-recharge so it tops up by itself when it drops below an amount.
  2. On the Plan tab, check that "Keep going with your Balance once you pass the cap" is on. It turns on automatically when you top up, unless you switched it off manually.

If both the monthly allowance and the balance reach zero, or that option is off, the AI pauses and your team keeps answering manually. Customer messages always keep coming in.

Example: 1,000 conversations a month

A business gets 1,000 conversations a month and its AI agent sends, on average, 6 replies per conversation (greeting, understanding what they want, answering questions, giving the price and closing or handing off).

1,000 × 6 = 6,000 AI replies a month.

  • Escala (about 6,000 included): just enough; it's worth keeping some balance in the wallet for busier months.
  • Crece (about 4,000): covers two thirds of the month; the other ~2,000 replies are paid from the wallet.
  • Conecta (about 2,000): covers one third; the other ~4,000 come from the wallet.

To decide, compare what you'd spend on top-ups with the price difference between plans (see Plans and what each includes): if you top up a lot every month, moving up a plan usually works out better.

How to use fewer replies without serving customers worse

  • Write clear instructions and keep a tidy knowledge base: the AI resolves things in fewer messages.
  • Ask for short answers and one question per message.
  • Hand off to a person when the case calls for it, instead of dragging out the conversation.
  • Check the day-by-day detail in Settings → AI usage to see which conversations or agents use the most.

Common problems

  • The AI stopped replying mid-month: the allowance ran out and there's no balance, or "Keep going with your Balance once you pass the cap" is off. Top up and check the option.
  • I used more than I estimated: check whether workflows or features (summaries, transcription) also use AI, and whether the agent replies with very long messages.
  • The agent's "Daily token budget" doesn't stop anything: that's right, that field no longer limits usage. The real control is the monthly allowance and the wallet.

Related articles

Was this helpful?

Related articles

Message us on WhatsApp