Token Cost Calculator

Token cost is the number of tokens processed divided by one million, multiplied by the model's per-million-token rate. Input (your prompt) and output (the response) are priced separately, so total cost is input tokens times the input rate plus output tokens times the output rate. This tool does that math for every model at once.

As a rule of thumb, one token is about four characters or roughly 0.75 words of English, so 1,000 tokens is around 750 words. This tool estimates tokens from your pasted text using the four-characters-per-token approximation. For an exact count, paste the text into your provider's tokenizer, then enter that number here directly.

Cost per million tokens is the standard way LLM APIs quote pricing. Rates range from a few cents per million for small open models to tens of dollars per million for frontier models, and output usually costs more than input. The table above lists the live input and output rate for every model so you can compare per-million prices side by side.

Sort the table by Price to see the cheapest models per million tokens; small open-weight models from Qwen, Llama, Mistral, and others are usually the lowest cost. But the cheapest model is not always the best value: sort by Your Cost to rank by the actual price for your text, and weigh quality against price for your task.

Prices come from the OpenRouter model catalog, which passes through each provider's rate at or near direct cost, and refresh daily. Quality scores come from Theozard. Always confirm against the provider's own pricing page before committing to production budgets, since rates change and new models ship regularly.