Chatbot pricing: flat-rate vs per-message vs per-resolution

AI chatbots are usually priced in one of three ways: a flat monthly fee, a charge per message or conversation, or a charge per resolved conversation. Flat-rate pricing is the easiest to budget for. Usage-based pricing can look cheaper at low volume, but it ties your bill to traffic you do not control. What matters most is understanding what is metered, and what happens when you reach the limit.

This guide explains each model in general terms, what actually drives chatbot cost, and how Chyt.ai's plans and AI credits work, with current prices in rupees.

What are the main chatbot pricing models?

Flat-rate pricing

Flat-rate pricing is a fixed monthly or annual fee for a defined bundle: a number of chatbots, a usage allowance, a set of channels and features. You know the bill before the month starts. The questions to ask are what the allowance covers and what happens when it runs out.

Per-message or per-conversation pricing

Per-message pricing charges for each message or conversation the bot handles, sometimes after a small included quota. It feels fair, because you pay for what you use. The difficulty is that usage is driven by your visitors. A successful campaign, a seasonal peak or a product issue can multiply the bill in a month when you least want a surprise.

Per-resolution pricing

Per-resolution pricing charges when the chatbot resolves a conversation without a human. It aligns cost with outcomes in principle. In practice, everything depends on how the vendor defines a resolution: a visitor who leaves without replying may or may not count, and you are relying on the vendor's measurement of its own success.

What actually drives chatbot cost?

Whatever the pricing model, a vendor's prices reflect the same underlying costs. Knowing them helps you read any pricing page.

• Model tokens. Every AI reply sends text to a language model and receives text back. Cost rises with the model chosen, the length of the conversation history, the amount of retrieved content included, and the length of the reply. A stronger model answering a complex question costs more than a fast model answering a simple one.
• Channels. A website widget typically involves no third-party message fee. Channels such as WhatsApp, SMS and telephony involve providers that may charge separately, usually on your own provider account.
• Seats. Some tools charge per agent or team member who uses the inbox or dashboard. Check whether your expected team size changes the price.
• Chatbots, content and storage. Plans often limit the number of bots and the volume of content you can index.
• Voice. Speech recognition and synthesis are metered by time, so voice is normally a separate allowance.
• Support and extras. White-labelling, custom domains, private model keys and priority support tend to sit on higher tiers.

How does Chyt.ai pricing work?

Chyt.ai uses flat monthly pricing in Indian rupees. It is never priced per message. Prices below exclude taxes; applicable GST and the final total are shown before you confirm payment.

Pro adds bring-your-own-key (BYOK) and private LLM support, API access and access to the wider set of supported channels, subject to availability and your own provider accounts. Enterprise adds white-label deployment and the agency portal. The pricing page has the full feature comparison.

How do AI credits work at Chyt.ai?

AI credits are Chyt.ai's single allowance for AI usage, shared across all chatbots in a workspace.

• What uses credits. Managed AI responses and platform AI features consume credits. Different models and request sizes consume different amounts, so a short answer from a fast model uses less than a long answer from a stronger one. Automatic model routing sends simple questions to a fast model and complex ones to a stronger model.
• What does not use credits. Messages handled entirely by deterministic flow nodes use no model tokens and no AI credits.
• When credits reset. Paid-plan credits reset each calendar month (UTC), including on annual billing. Trial credits cover the whole trial and do not renew.
• What happens at zero. Managed AI requests pause until credits are available. Configured deterministic flows can continue while the plan is active.
• BYOK. On Pro and Enterprise you can connect your own model provider account. Your provider then bills you for that inference directly, while platform AI features may still use workspace credits.
• Voice is separate. Browser voice chat has its own time allowance, tracked separately from AI credits. AI credits still apply to the AI answers within a voice conversation.

There are no fixed monthly conversation or token quotas. The model settings page shows estimated AI conversation figures based on remaining credits and stated assumptions; they are estimates, not promised counts. The help guide on plans, AI credits and voice allowances explains how to read the usage card.

Why do unlimited deterministic conversations matter for budgeting?

Every active Chyt.ai plan includes unlimited deterministic flow conversations. A deterministic flow is a guided conversation built from fixed steps: menus, qualification questions, routing and closing messages. Because no model is called, these conversations do not draw on your AI credits however many visitors use them.

That changes the budgeting question. Many chats follow a few predictable paths: choosing a department, qualifying an enquiry, checking basic eligibility. If those run as flows, your AI credits are reserved for the open questions where generated answers earn their cost. A traffic spike on the scripted paths does not change the bill or exhaust the allowance.

It also gives you a safety net. If AI credits run out late in the month, the guided part of your chatbot keeps working. Our comparison of deterministic and AI chatbots explains how to decide which parts of a conversation belong in a flow.

How much does a chatbot cost in India?

It depends on the pricing model and what is metered, so an honest answer is a method rather than a single number. Estimate your monthly conversations, split them into predictable paths and open questions, and ask each vendor what both would cost in a busy month as well as a normal one. Check whether prices are in rupees or a foreign currency, and whether GST is included.

As a concrete reference, Chyt.ai's paid plans start at ₹3,999 per month plus GST for five chatbots, or ₹3,199 per month when billed annually, and every plan starts with a free 14-day trial.

What should you ask a chatbot vendor about pricing?

1. What exactly is metered: messages, conversations, resolutions, credits, tokens or seats?
2. What happens when I reach the limit: does the bot stop, slow down, or bill overage automatically?
3. Can I set a spending cap?
4. How is a "conversation" or "resolution" defined, and can I audit the count?
5. Do scripted or button-based conversations count against my AI allowance?
6. Are channel provider fees, such as messaging or telephony charges, included or separate?
7. Is pricing per seat, and how many team members are included?
8. Which features sit on higher tiers: custom branding, API access, own model keys, white-label?
9. Can I use my own model provider account, and what does the platform still charge for?
10. Are prices in rupees, and are taxes included?
11. What happens to my chatbots and data if I downgrade or cancel?

Next steps

Compare the plans on the Chyt.ai pricing page, or start a free trial to see how far the included credits go with your own content. To understand the usage card once you are in, read the help guide on plans, AI credits and voice allowances.