AI Token & Cost Calculator

What This Calculator Measures

The AI Token & Cost Calculator turns a prompt (and optional completion) into two numbers that actually matter when you're building with an LLM API: how many tokens it uses, and what that costs in dollars, model by model.

How Tokens Work

Language models don't read text character-by-character or word-by-word — they break it into tokens, chunks of text produced by a byte-pair-encoding (BPE) algorithm trained on huge amounts of text. A token is often a whole short word, a word fragment, or a punctuation mark. As a rough rule of thumb, 1 token is about 4 characters or 0.75 words of English text, but this varies a lot by language and content type (code and non-English text often use more tokens per character).

Why Accuracy Differs by Provider

• OpenAI models (GPT-5.x, GPT-4.1, o-series) all use OpenAI's o200k_base BPE vocabulary. That vocabulary is public, so this calculator runs the exact same tokenizer algorithm entirely in your browser — no network call, no API key, and the count matches what OpenAI's own API would report.
• Anthropic Claude and Google Gemini don't publish an offline tokenizer. The only way to get an exact count is to call the provider's own token-counting endpoint (Anthropic's /v1/messages/count_tokens, Google's countTokens). If you add your own API key in the calculator, we forward it to that endpoint for a single request (never stored) and show you the real number. Without a key, you'll see a clearly labeled estimate instead.

The Cost Formula

input cost  = (input tokens / 1,000,000)  × input price per million tokens
output cost = (output tokens / 1,000,000) × output price per million tokens
total cost  = input cost + output cost

If a model supports prompt caching and you mark the input as a cache hit, the calculator swaps in the discounted cached-input rate instead of the standard input price.

Worked Example

Say your prompt tokenizes to 1,200 tokens and the model replies with 400 tokens, using a model priced at $2.50 input / $15.00 output per million tokens:

• Input cost: 1,200 / 1,000,000 × $2.50 = $0.003
• Output cost: 400 / 1,000,000 × $15.00 = $0.006
• Total: $0.009 per call

Multiply that by your expected calls per day to see your real daily spend.

Common Mistakes

• Forgetting output tokens can dominate cost. Output pricing is usually 4-6x higher than input pricing, so a short prompt with a long answer can cost more than a long prompt with a short answer.
• Assuming all providers tokenize the same way. The same sentence can produce different token counts on GPT vs. Claude vs. Gemini — always price with the actual model's count, not a generic estimate, when the difference matters to your budget.
• Ignoring system prompts and tool definitions. If you're using function calling or a long system prompt, those tokens count too and should be included in what you paste into the calculator.

Frequently Asked Questions