How much does a chatbot cost per month? (GPT, Claude, Gemini)
Thinking about adding a chatbot to your site, store, or team tool? Before you sign with a provider or send a quote to a client, check what it really costs per month. No guessing here: you type how many questions per day, how long the question, how long the reply, and you get a concrete number.
The calculator shows daily, monthly, and yearly cost at once for the 7 most popular bots (GPT-4o, GPT-4o mini, Claude Sonnet, Claude Haiku, Claude Opus, Gemini Pro, Gemini Flash). The cheapest one is highlighted green, so you instantly see where the savings are.
Prices are typed in by hand from the official pricing pages of OpenAI, Anthropic, and Google, with the last verification date shown next to each model. You also get a USD/EUR toggle (set the rate yourself) and a slider for the discount on repeating parts of the input, which can be as much as 90% off.
How to use it
- Type how many questions per day, meaning how many times someone hits Send in the chat (one click is one question).
- Type the average question length in tokens. A token is a chunk of text the bot sees as one unit. 1 English word ≈ 1.3 tokens.
- Type the average reply length the same way, in tokens.
- The table at the bottom shows daily, monthly, and yearly cost for each of the 7 bots. The cheapest is highlighted green, plus how much you save versus the most expensive.
- The % of questions with the repeated-parts discount slider: if you send the same instruction every time (*you are a store assistant, reply politely…*), the bot gives a discount on it. The slider shows how much you save.
- USD/EUR toggle: set the rate manually (e.g. 0.92 EUR per 1 USD).
- "Top 5 only" view hides the most expensive models so the table stays clean.
When this is useful
Six typical situations where the calculator gives you a concrete answer instead of a guess:
- Quoting a client project. The client asks: *"what will a chatbot for 5,000 customers a day cost me?"*. You enter the numbers and show a concrete monthly figure instead of guessing. The client sees you actually know the math.
- Choosing which bot to use. Claude Sonnet or GPT-4o for customer support? The difference looks tiny (a few cents per question), but multiplied by 365 days and 100 customers per day it adds up to thousands of dollars a year. You see it immediately, no mental math.
- Checking if migrating to a cheaper bot pays off. You're using an expensive one (e.g. Claude Opus), but for 80% of your tasks a cheaper one is enough (Claude Haiku). The calculator shows it can be a dozen times cheaper at the same quality on simple questions. The decision makes itself.
- Checking if the repeated-parts discount is worth it. You have a long instruction for the bot (5,000 tokens: *"respond like a financial advisor, never promise returns…"*) that repeats with every question. The slider shows how much you save by enabling the discount and whether it's worth the setup.
- Startup budgeting. You have $100 a month for the bot. How many questions per day can you afford on cheaper Gemini Flash versus pricier GPT-4o? Type the budget, play with numbers, see the ceiling.
- Comparing all 7 models before launch. Instead of reading seven pricing pages and pasting them into a spreadsheet, you get a side-by-side table instantly. Decision based on data, not gut feeling.