Recommended tools
How this calculator works
Daily cost is calculated by taking your daily input tokens divided by one million and multiplying by the provider's input price per million tokens, then doing the same for output tokens with the output price. Monthly cost multiplies the daily cost by your number of active usage days per month. The calculator accounts for the fact that output tokens typically cost 3-5 times more than input tokens due to the compute intensity of generating text versus processing it. Caching discounts and batch pricing — which can reduce costs by up to 50% in some cases — are noted as potential savings beyond the base calculation.
Useful scenarios
- A developer comparing OpenAI GPT-4o vs Anthropic Claude for a chatbot handling 100K input and 20K output tokens daily.
- A solo founder building an AI writing assistant — estimating monthly API bills at different usage levels to decide between providers.
- A content creator using AI for batch processing — comparing DeepSeek V3 vs Gemini 2.5 Pro for 500K daily tokens on heavy processing days.
FAQ
Why are input and output tokens priced differently?
Output tokens cost more because generating text requires more compute than processing input. Most providers charge 3-5x more for output tokens. For chat-heavy workloads, output costs usually dominate.
How many tokens does a typical API call use?
A short chat message uses 200-500 tokens. A code review with context uses 4,000-8,000 tokens. Document analysis with a 50-page PDF uses 15,000-30,000 input tokens. Adjust your daily token estimates based on your actual use case.
Should I consider prompt caching?
Yes. If you send the same system prompt repeatedly, caching can reduce input costs by up to 50% on some providers. This calculator does not model caching — your actual costs may be lower if you implement it.