Enter your token volume and instantly compare what GPT-4o, Claude, and Gemini would cost — per call and per month. Editable rates, sorted cheapest first. Free, private, no sign-up.
Not sure how many tokens your prompt is? Count them first →
| Model | $/1M in | $/1M out | Per call | Per month |
|---|---|---|---|---|
| Gemini Flashcheapest | $0.00030 | $0.3000 | ||
| GPT-4o mini | $0.00045 | $0.4500 | ||
| Claude Haiku | $0.00280 | $2.80 | ||
| Gemini Pro | $0.00375 | $3.75 | ||
| GPT-4o | $0.00750 | $7.50 | ||
| Claude Sonnet | $0.0105 | $10.50 | ||
| Claude Opus | $0.0525 | $52.50 |
Rates are editable, approximate 2026 figures ($ per 1M tokens) — always confirm current pricing with the provider. Calculation is an estimate.
Shorter, sharper prompts cut token cost. Trim yours
Enter your input tokens per call, output tokens per call, and calls per month. The tool multiplies those by each model's price per million tokens and shows cost per call and per month, sorted cheapest first. All rates are editable so you can match current provider pricing exactly.
They're approximate 2026 defaults and fully editable. AI pricing changes often and differs for input vs output tokens, so always confirm the current rate on the provider's pricing page — then adjust the fields here for an exact estimate.
Input tokens are what you send (your prompt + context); output tokens are what the model generates. Output is usually priced higher than input, which is why the calculator separates them — long responses can dominate your bill.
Use a smaller/cheaper model where quality allows, trim unnecessary context from prompts, cap output length, cache or reuse results, and batch work. Tighter prompts reduce both input and output tokens — our token counter and prompt improver help with that.
Yes, completely free with no sign-up. The calculator runs entirely in your browser — nothing is uploaded.
Tighter prompts use fewer tokens and get better results. Grade yours, count its tokens, or start from 25,000+ optimized prompts.