Paste any text and instantly see the estimated token count and cost for popular AI models. Everything runs in your browser.
| Model | Est. Tokens | Input Cost | Price / 1M tokens |
|---|
How Token Counting Works
AI models don’t read text the way humans do. They break text into tokens — chunks that are roughly 4 characters or 0.75 words in English. The exact count varies by model and tokenizer, but the ratio is consistent enough for cost estimation.
Rules of thumb:
- 1 token ≈ 4 characters in English
- 1 token ≈ 0.75 words
- 100 tokens ≈ 75 words
- Code typically uses more tokens per line than prose
Why Token Counts Matter
You pay for AI APIs per token. Both input (what you send) and output (what you get back) count. Knowing your token usage helps you:
- Estimate API costs before committing to a model
- Stay within context window limits
- Optimize prompts to reduce costs
- Compare pricing across providers
A Note on Accuracy
This tool uses a character-based estimation (1 token per 4 characters) with model-specific adjustments. For exact counts, you’d need each model’s actual tokenizer. But for cost estimation and planning, this gets you within 5-10% of the real number — close enough to make informed decisions.