Skip to main content

Do you need your own OpenAI key for an AI chatbot? Included vs bring-your-own, explained

Rosen Hristov 4 min read

Some AI chatbots make you open an OpenAI account, generate a key, paste it in, then pay a second token bill that can climb without warning. Emporiqa includes the AI model cost inside one per-conversation price. There is no key to manage and no second invoice to track.

If you have shopped for an AI chatbot for your store, you may have hit a step that stopped you cold. The tool asks you to open an account with OpenAI, generate something called an API key, and paste that key into the chatbot settings. Then, on top of whatever you pay the chatbot company, you start getting a second bill from OpenAI for the AI usage. This setup has a name. It is called bring your own key, often shortened to BYOK.

For a non-technical store owner, that is two problems at once. You have to set up and manage a developer account you never asked for, and you now have two bills to watch instead of one, with the AI bill being the one that moves.

What bring-your-own-key means

The AI behind a chatbot is rented from a provider like OpenAI. With BYOK, the chatbot company does not pay that provider. You do. The chatbot is the software layer; your own OpenAI key is what pays for the thinking underneath it. So your monthly cost splits in two: the fee for the chatbot tool, plus whatever OpenAI charges your key for the month.

That second part is billed by the token. A token is a small piece of text, roughly a few characters. Every question a shopper types and every answer the chatbot gives is counted in tokens. A quiet week costs little. A sale, a product launch, or one long back-and-forth chat costs more. You only find out the total after the month is over.

Why the token bill is hard to predict

The trouble with a token bill is that it follows traffic you do not fully control. Consider what moves it:

  • A spike in visitors after an email blast or an ad.
  • Longer conversations, because the chatbot is doing its job well.
  • A new, larger catalog that gives the chatbot more to read.
  • Off-hours chats at 2am that you are not even awake to see.

None of those are bad for your business. Some are what you wanted. But each one pushes the token bill up, and there is no clean cap unless you go set one yourself inside the provider dashboard. Many owners do not know that dashboard exists until the bill surprises them.

How Emporiqa handles it instead

Emporiqa includes the AI model cost inside the price you already pay. The pricing is a flat price per conversation, plus a charge per AI-drafted email reply you send; the rates are on the pricing page. The conversation price covers the chatbot and the AI running behind it together. There is no OpenAI account to open and no key to manage. You won't get a second invoice from a provider to track.

Because the AI is managed centrally on our side, you also skip the token math. You do not estimate tokens, you do not guess at a monthly range, and you do not log into a provider console to set a limit. One conversation, one known price.

What the numbers look like

Here is how the pricing works, with nothing hidden behind a key (the rates are on the pricing page):

  1. No monthly base. No subscription floor.
  2. A flat price for each conversation a shopper has with the chat.
  3. A charge for each AI-drafted email reply you send, if you connect your shop's mailbox.
  4. Free to start, with signup credit added when you create a store.
  5. No card required to sign up. You add one only when the credit runs out.
  6. A monthly limit you set, which you can raise, lower or remove yourself from the billing dashboard.

That limit matters. It is your safety rail set in plain dollars, not in tokens you have to translate. If a busy month worries you, you set the ceiling and the spend stops there, and so does the chat until you raise it. You change it from the billing dashboard, not a provider console.

What you get for that price

Emporiqa is an AI chatbot that sells before the purchase and serves after it on your store. It recommends products from your own catalog, by text or from a photo a shopper uploads, answers questions about them, suggests variations, and adds products to the cart, in their own language. Handoff to a person is built in. Order status works once your store exposes order lookup, which the native plugins set up, and return and cancellation requests are store rules you switch on when you want them. The chat passes return and cancellation requests to your team by email; it does not refund or cancel by itself. All of that runs on the AI cost that is already inside the per-conversation price.

Install is just as plain. You add the official plugin from your platform's marketplace (WooCommerce, Magento, PrestaShop, Drupal Commerce or Sylius), or install it with Composer on Shopware 6, then click Connect in the plugin admin. The basic setup needs no API key and no developer.

One bill, no key

If the idea of opening a developer account and babysitting a second bill is what has held you back, that whole step is gone here. You connect the plugin and the AI is handled. The cost you see is the cost you pay.

Ready to skip the key and the token math? Create a free Emporiqa account. No card needed. Or watch it work first at demo.emporiqa.com.

If you run WooCommerce, see how to add a WooCommerce chatbot with no API key.

Frequently asked questions

Do I need my own OpenAI API key to use Emporiqa?

No. Emporiqa runs the AI for you. There is no OpenAI account to open and no key to paste. The model cost is already inside the price you pay per conversation and per AI-drafted email reply you send, so you set it up and connect, nothing more.

Who pays for the AI when a shopper chats?

Emporiqa does, and it is already covered in your per-conversation price. You pay each time a shopper has a conversation and for each AI-drafted email reply you send; that amount includes the model running behind it. There is no separate provider bill arriving on its own.

What is a token bill and why is it hard to predict?

A token bill is what an AI provider charges per chunk of text processed. A busy day or a long chat uses more, so the number moves around and can rise without warning. Emporiqa removes that math by charging a flat price per conversation and per AI-drafted email reply you send, model cost included.

Rosen Hristov, founder of Emporiqa

Rosen Hristov

Founder of Emporiqa

Connect on LinkedIn

Share this article

Related Articles

Try it on your store

Connect your products and see the chat answer shopper questions from your own catalog.

Your dashboard shows what each conversation led to.

  • Free to start
  • Pay per use, no seat fees
  • No card required