Token Calculator
A token calculator estimates how many tokens your text uses in a large language model and what those tokens cost. Paste a prompt or document to see characters, words and estimated tokens, then compare the input and output cost across OpenAI, Anthropic and Google models. Free, no sign-up. Prices checked 1 October 2026.
- Free, no sign-up
- 10 models priced
- Input and output cost
Count Tokens and Estimate Cost
Characters
186
Words
35
Estimated tokens
47
Estimate based on OpenAI's rules of thumb for English text. Exact counts differ by model and language.
Usage
Your text is treated as the input of each request. Reasoning models also bill hidden reasoning tokens as output, so add headroom for them.
Cost by model (cheapest first)
| Model | Input cost | Output cost | Per request | Per month |
|---|---|---|---|---|
| GPT-6 Luna ($0.1/$0.5 per 1M) | $0.000005 | $0.000250 | $0.000255 | $2.55 |
| Gemini 3.5 Flash-Lite ($0.3/$2.5 per 1M) | $0.000014 | $0.001250 | $0.001264 | $12.64 |
| Claude Haiku 4.5 ($1/$5 per 1M) | $0.000047 | $0.002500 | $0.002547 | $25.47 |
| Gemini 3.5 Flash ($1.5/$9 per 1M) | $0.000070 | $0.004500 | $0.004571 | $45.71 |
| GPT-6.1 Sol ($2/$10 per 1M) | $0.000094 | $0.005000 | $0.005094 | $50.94 |
| Claude Sonnet 5.5 ($2/$10 per 1M) | $0.000094 | $0.005000 | $0.005094 | $50.94 |
| Gemini 3.1 Pro (preview) ($2/$12 per 1M) | $0.000094 | $0.006000 | $0.006094 | $60.94 |
| Claude Opus 5.5 ($4/$20 per 1M) | $0.000188 | $0.0100 | $0.0102 | $101.88 |
| GPT-6 Astra ($10/$50 per 1M) | $0.000470 | $0.0250 | $0.0255 | $254.70 |
| Claude Fable 5.1 ($10/$50 per 1M) | $0.000470 | $0.0250 | $0.0255 | $254.70 |
Standard API prices per 1M tokens, checked 1 October 2026: OpenAI, Anthropic, Google. Caching and batch discounts are not applied; the LLM cost calculator models them.
What is a token in an LLM?
A token is a chunk of text that a language model processes as one unit. It can be a whole word, part of a word, a punctuation mark or a space. Common English words are often a single token, while rare words, names, code and numbers are split into several.
Providers charge separately for input tokens (your prompt, instructions and any documents you include) and output tokens (the model's reply). Output tokens cost more, so a long answer can cost more than a long prompt. Reasoning models also generate internal reasoning tokens that are billed as output even though you do not see them.
Each model family uses its own tokenizer, so the same text produces slightly different counts on OpenAI, Anthropic and Google models. Non-English text often needs noticeably more tokens per word.
How many tokens is a word?
| 1 token | ≈ 4 characters of English text |
| 1 token | ≈ ¾ of an English word |
| 100 tokens | ≈ 75 words |
| 1,000 words | ≈ 1,333 tokens |
| 1 page (≈ 500 words) | ≈ 667 tokens |
English text. Source: OpenAI Help Center.
How do you calculate token cost?
Cost per request = (input tokens ÷ 1,000,000 × input price) + (output tokens ÷ 1,000,000 × output price). Multiply by requests per month for a monthly bill. For an agent, multiply again by the number of model calls each task makes.
Example: a 1,500-token prompt with a 500-token reply on a model priced at $2 input and $10 output per million tokens costs $0.003 + $0.005 = $0.008 per request, or $80 for 10,000 requests a month.
How can you use fewer tokens?
Trim repeated instructions, send only the relevant parts of long documents, cap reply length, and use prompt caching for content you send on every request. Routing simple requests to a smaller model usually saves more than any prompt edit. If token spend is becoming a real line item, our AI development team can help design for cost from the start.
Frequently asked questions
How do I calculate tokens from words?
For English text, divide the word count by 0.75, or multiply it by about 1.33. OpenAI's guidance is that one token is roughly four characters or three-quarters of a word, so 100 tokens is about 75 words. Code, numbers and other languages usually use more tokens per word, so treat the result as an estimate and check exact counts with your provider's tokenizer.
How much do 1,000 tokens cost?
It depends on the model and whether the tokens are input or output. At prices checked on 1 October 2026, 1,000 input tokens cost $0.0001 on GPT-6 Luna, $0.002 on GPT-6.1 Sol or Claude Sonnet 5.5, and $0.01 on GPT-6 Astra or Claude Fable 5.1. Output tokens cost about five times more than input tokens on these models.
Do all models count tokens the same way?
No. Each model family uses its own tokenizer, so the same text can produce different token counts on OpenAI, Anthropic and Google models, and different languages tokenize very differently. This calculator uses the common English rule of thumb. For billing-grade accuracy, use the provider's own token-counting tool or the usage figures returned by the API.
Related Tools
LLM Cost Calculator
Turn token counts into monthly cost per request, per agent task and per user, with a 12-month projection.
Try it free CalculatorLLM Observability Cost Calculator
Compare what Langfuse, LangSmith and Datadog would cost for your trace volume, seats and retention.
Try it free CalculatorAI ROI Calculator
Calculate your AI investment return with detailed 3-year projections, sensitivity analysis, and payback period.
Try it freeTrack token cost per model, agent and user
This calculator estimates tokens before you build. Perimattic AI Suite measures them in production, with token cost per model, agent, feature and user.
Or email sales@perimattic.com
Last updated 19 October 2026