—
est. tokens—words
—characters
—% of 128k context
Cost to send this text as input
Your token count × each model's input price, ranked cheapest first. Output billed separately.
| Model | Input $/1M | This text costs |
|---|
⚠️ Token count is an estimate (~4 characters per token). Real tokenizers split on sub-words, so code, numbers, emoji and non-English text vary by a few percent. Prices are reference (July 2026). For exact counts use the provider tokenizer (e.g. tiktoken for OpenAI). · Report outdated price →
How token counting works
Language models don't read characters or words — they read tokens, chunks of roughly 4 characters or 0.75 words of English. So 1,000 tokens ≈ 750 words. You're billed per token on both what you send (input) and what the model writes back (output), which is why a long system prompt you resend every call quietly adds up. Paste your real text above to see its size and cost before you ship it.
Next steps
Now estimate a whole workload: the AI API cost calculator, the LLM token cost calculator (input + output), or the AI app cost estimator. Resending the same context each call? Prompt caching cuts it.