Token Economics Calculator
Paste any text or prompt. Estimate token count, configure output length and caching, then compare real API costs across — data refreshes every 24 h from official sources.
Output Multiplier
Estimated output tokens relative to input
Prompt Caching
Apply provider cache-read rates to input tokens. Anthropic, OpenAI, and Google all offer ~90% off cached input.
Request Summary
Paste text above to see costs
Sort by provider · Click cost headers to re-sort · Non-cached rates
Per 1M tokens · Standard rates
How tokenization is estimated
This calculator divides character count by 4 — the well-known BPE heuristic used by GPT, Claude, and Gemini tokenizers for typical English input. For mixed-language text or code-heavy prompts, actual token counts may differ by ±10–20%. For exact counts, use OpenAI's Tokenizer or Anthropic's token counter API.
Was this page helpful?
Sign in to cast your vote
Discussion
Sign in to share your feedback and join the discussion.