Power users

Prompt Cost Estimator

Calculate what one prompt, repeated test, or production workflow costs to run.

Pricing table last updated: 2026-04-26

Interactive tool

Estimate prompt cost

Adjust the inputs and the result updates instantly.

Cost / run$0.0415
Total prompt cost$41.50

GPT-5.5: $5.00 input / $30.00 output per 1M tokens.

How this prompt cost estimator works

A prompt cost estimator helps power users understand the price of a specific prompt. This is useful when you are testing prompt variations, running evaluations, building agents, or designing workflows that call a model multiple times. A prompt that costs almost nothing once can become expensive when repeated thousands of times in testing or production.

Choose a model, enter the prompt input tokens, expected output tokens, and number of runs. The calculator shows the cost per run and total cost. This makes it easier to compare prompt versions. A shorter prompt with the same quality may save money every time it runs. A longer prompt may still be worth it if it improves accuracy enough to reduce retries or human review.

For agentic workflows, remember that one user action may trigger multiple prompts. Planning, tool calls, retrieval, validation, and final response generation can each add tokens. If you are building an AI agent or automation, estimate the whole chain rather than only the final answer. That gives a more realistic view of production cost.

The estimator uses text token prices and does not include images, audio, search grounding, tool fees, storage, or provider discounts. For prompt iteration and cost awareness, the baseline is still valuable. It helps teams make cost a normal part of prompt design instead of a surprise at billing time.

Examples

Estimate the cost of running an eval prompt 5,000 times.

Compare a short prompt and a long prompt before production rollout.

Model an agent workflow where each user task triggers several runs.

FAQ

Are these AI costs exact?

They are estimates based on public token prices. Your bill can change with cached tokens, batch discounts, image or audio usage, taxes, provider credits, and model-specific rules.

Which currency does the calculator use?

All calculators use USD by default because major AI providers publish API pricing in USD.

Why multiply by runs?

Testing, evals, batch jobs, and production workflows repeat prompts. Total cost depends on both cost per run and how often the prompt runs.