Skip to content

AI Chat Cost Calculator

Text ToolsAI Writing Assistants
Chat volume

Prompt, system instructions and any chat history sent with the message.

The length of the model's reply.

Model pricing
Monthly cost (USD) —
Cost per message (USD)
—
Daily cost (USD)
—
Yearly cost (USD)
—
Tokens per month
—
Share of cost from output
—
Monthly cost: input vs output tokens
The chart appears with your result

Chat API cost is tokens multiplied by the price per million tokens, with input and output priced separately. A message with 500 input tokens at $2.50 per million and 300 output tokens at $10 per million costs $0.00425; 1,000 such messages a day for 30 days come to $127.50 a month.

About this tool

Teams putting a chatbot, support assistant or internal helper on a paid language-model API need a budget before traffic arrives. Providers bill by the token, and almost all of them charge more for the tokens the model writes than for the tokens it reads. This calculator takes your chat volume, the average token count on each side of a message and the two prices from your provider's price list, then shows the cost per message, per day, per month and per year, plus how much of the bill comes from output. A donut chart splits the monthly total between input and output. It uses the prices you type, so it stays correct when a provider changes its list. It does not model prompt caching discounts, batch pricing, image or audio tokens, or per-seat subscription plans.

How to use it

  1. Describe the chat volume

    Set messages per user per day, use the stepper for the number of users, and drag the slider to the number of active days in a month.

  2. Enter token counts

    Type the average input tokens (prompt plus resent history) and output tokens (the reply) for one message.

  3. Add the model's prices

    Copy the input and output price per million tokens from the provider's pricing page.

  4. Read the breakdown

    The monthly figure leads; per-message, daily and yearly cost, total tokens and the output share follow, with the input/output split in the donut chart.

Examples

1,000 chats a day, $2.50 / $10 model

Messages per user per day
1000
Users
1
Active days per month
30
Input tokens per message
500
Output tokens per message
300
Input price per 1M tokens
2.5
Output price per 1M tokens
10

Result Monthly cost (USD): 127.50
Cost per message (USD): 0.00425
Daily cost (USD): 4.25
Yearly cost (USD): 1,530.00
Tokens per month: 24,000,000
Share of cost from output: 70.6%

  1. Input per message = 500 × $2.5 ÷ 1,000,000 = $0.00125
  2. Output per message = 300 × $10 ÷ 1,000,000 = $0.003
  3. Per message = $0.00425; messages per day = 1000 × 1 = 1000
  4. Monthly = $0.00425 × 1000 × 30 days = $127.5

Output tokens cost four times as much here, so the 300-token replies make up about 71% of the bill despite being the smaller count.

5 users, 200 messages each, 22 workdays

Messages per user per day
200
Users
5
Active days per month
22
Input tokens per message
2000
Output tokens per message
500
Input price per 1M tokens
3
Output price per 1M tokens
15

Result Monthly cost (USD): 297.00
Cost per message (USD): 0.0135
Daily cost (USD): 13.50
Yearly cost (USD): 3,564.00
Tokens per month: 55,000,000
Share of cost from output: 55.6%

  1. Input per message = 2000 × $3 ÷ 1,000,000 = $0.006
  2. Output per message = 500 × $15 ÷ 1,000,000 = $0.0075
  3. Per message = $0.0135; messages per day = 200 × 5 = 1000
  4. Monthly = $0.0135 × 1000 × 22 days = $297

Each message costs $0.0135; 1,000 messages a day over 22 working days comes to $297.

3,000 messages on a $0.15 / $0.60 model

Messages per user per day
100
Users
1
Active days per month
30
Input tokens per message
1000
Output tokens per message
1000
Input price per 1M tokens
0.15
Output price per 1M tokens
0.6

Result Monthly cost (USD): 2.25
Cost per message (USD): 0.00075
Daily cost (USD): 0.08
Yearly cost (USD): 27.00
Tokens per month: 6,000,000
Share of cost from output: 80%

  1. Input per message = 1000 × $0.15 ÷ 1,000,000 = $0.00015
  2. Output per message = 1000 × $0.6 ÷ 1,000,000 = $0.0006
  3. Per message = $0.00075; messages per day = 100 × 1 = 100
  4. Monthly = $0.00075 × 100 × 30 days = $2.25

Six million tokens a month cost $2.25 on a low-priced model, with output still 80% of the spend.

How it is calculated

Cost per message = (Tin × Pin + Tout × Pout) ÷ 1,000,000; Monthly cost = cost per message × messages per day × users × days

Tin
input tokens per message
Tout
output tokens per message
Pin
price per million input tokens, USD
Pout
price per million output tokens, USD

API price lists quote dollars per million tokens, so each side of a message is its token count times its price, divided by one million. The two sides are added to give the cost of one message, which is multiplied by the daily message count across all users and by the active days. Yearly cost is twelve months at that level. The output share is the output cost divided by the cost per message.

When not to use it

  • Flat-fee consumer subscriptions are billed per seat, not per token.
  • Workloads relying on cached prompts or batch discounts will cost less than shown.
  • Image, audio or video inputs use separate token prices this tool does not include.

Common mistakes

  • Counting only the new user message as input, when the full chat history is resent each turn.
  • Entering the price per 1,000 tokens in a field that expects the price per million.
  • Mixing up the input and output prices, which can understate the bill several times over.
  • Treating words as tokens; English text runs about 1.3 tokens per word.

Frequently asked questions

Why do output tokens cost more than input tokens?

Generating a token requires a full forward pass of the model for each new token, one after another, while input tokens are processed in parallel. Providers pass that difference on, so output prices are commonly three to five times the input price.

How many tokens is a typical chat message?

A short question is often 20 to 60 tokens, but the input of a real chat request also contains the system prompt and the earlier turns. After ten turns, input per message can easily be 2,000 tokens or more, even when each reply is only 300.

How can I convert words to tokens?

For English prose, one token averages about four characters or roughly three quarters of a word, so 1,000 words is around 1,300 tokens. Code, numbers and non-Latin scripts usually take more tokens per word. Your provider's tokenizer gives the exact count.

Does the chat history really count again on every turn?

Yes. Chat APIs are stateless, so each request resends the conversation so far as input. That is why long conversations get more expensive per message, and why trimming or summarising history is one of the most effective ways to lower spend.

Where do I find the prices to enter?

Each provider publishes a pricing page listing input and output prices per million tokens for every model. Enter the figures for the exact model and tier you call, since prices differ widely between a flagship model and a small one from the same company.

Why is my real invoice different from the estimate?

Averages hide spread: a few very long conversations can dominate spend. Invoices may also include cached-input discounts, batch discounts, tool-call tokens, reasoning tokens billed as output, or taxes. Compare a week of usage logs against the estimate and adjust the token averages.

Cite this tool

Using this tool in an assignment, article or report? Copy a citation, or a link for your website.

APA

Tools Directory Editorial Team. (2026, October 25). AI Chat Cost Calculator. Tools Directory. Retrieved October 26, 2026, from https://tools.milunir.com/text-tools/ai-writing-assistants/ai-chat-cost-calculator/

MLA

Tools Directory Editorial Team. “AI Chat Cost Calculator.” Tools Directory, 25 Oct. 2026, https://tools.milunir.com/text-tools/ai-writing-assistants/ai-chat-cost-calculator/. Accessed 26 Oct. 2026.

Chicago

Tools Directory Editorial Team. “AI Chat Cost Calculator.” Tools Directory. Last modified October 25, 2026. Accessed October 26, 2026. https://tools.milunir.com/text-tools/ai-writing-assistants/ai-chat-cost-calculator/.

Link<a href="https://tools.milunir.com/text-tools/ai-writing-assistants/ai-chat-cost-calculator/">AI Chat Cost Calculator – Tools Directory</a>