AI token calculator.

One million input tokens costs between $0.20 (Llama 4 Maverick) and $15 (Claude Opus 4), and output tokens cost 2-5x more per million. GPT-4o is $2.50/M input and $10/M output; Gemini 2.5 Flash is $0.30/M input. The token calculator prices any input/output mix across 12 models in one view.

One million input tokens costs between $0.20 (Llama 4 Maverick) and $15 (Claude Opus 4), and output tokens cost 2-5x more per million. GPT-4o is $2.50/M input and $10/M output; Gemini 2.5 Flash is $0.30/M input. The token calculator prices any input/output mix across 12 models in one view.

Enter the number of input and output tokens you expect to use, pick the models you want to compare, and the calculator instantly shows the cost in USD for every model, per request, per 100K tokens, or extrapolated to a monthly volume. It is free and requires no sign-up.

Enter the number of input and output tokens you expect to use, pick the models you want to compare, and the calculator instantly shows the cost in USD for every model, per request, per 100K tokens, or extrapolated to a monthly volume. It is free and requires no sign-up.

It depends on the model and the input/output mix. For example, 100K input tokens cost about $0.25 on GPT-4o, $0.30 on Claude Sonnet 4, and $0.03 on Gemini 2.5 Flash. Output tokens cost roughly 3-5x more per token than input tokens on most models.

It depends on the model and the input/output mix. For example, 100K input tokens cost about $0.25 on GPT-4o, $0.30 on Claude Sonnet 4, and $0.03 on Gemini 2.5 Flash. Output tokens cost roughly 3-5x more per token than input tokens on most models.

A token is the smallest unit of text a model reads or writes. One token is roughly 0.75 words in English, so 100K tokens is about 75,000 words, the length of a short book. Providers bill per token, both for the prompt you send (input) and the text the model generates (output).

A token is the smallest unit of text a model reads or writes. One token is roughly 0.75 words in English, so 100K tokens is about 75,000 words, the length of a short book. Providers bill per token, both for the prompt you send (input) and the text the model generates (output).

Generating text requires the model to run its full forward pass once per output token, while input tokens are mostly processed in parallel. That is why output prices are typically 3-5x higher than input prices on the same model.

Generating text requires the model to run its full forward pass once per output token, while input tokens are mostly processed in parallel. That is why output prices are typically 3-5x higher than input prices on the same model.

Payle is building financing for AI spend, including usage-based costs. With an annual Payle plan you pay $0 today and spread the estimated monthly API cost over 3 equal installments at 0% APR, so heavy token usage does not require a large upfront budget.

Payle is building financing for AI spend, including usage-based costs. With an annual Payle plan you pay $0 today and spread the estimated monthly API cost over 3 equal installments at 0% APR, so heavy token usage does not require a large upfront budget.