Batch summaries
100k input tokens and 50k output tokens for a content batch.
How this token cost calculator works
A token cost calculator is the fastest way to translate model pricing into a real number. AI providers usually publish prices per one million tokens, but developers often think in individual prompts, documents, conversations, or batch jobs. This creates a gap between the pricing table and the decision you need to make. By entering the number of input tokens and output tokens, you can estimate what one unit of work will cost.
Input and output tokens usually have different prices. Output tokens are often more expensive because generation requires more inference work from the provider. That means two prompts with the same total token count can have different costs depending on whether most tokens are sent to the model or generated by it. For example, a classification job may have many input tokens but a tiny output, while a writing assistant may generate long answers and become more output-heavy.
This calculator is useful for prompt engineering, batch processing, internal tools, and usage-based SaaS planning. You can paste token counts from your logs, estimate a workflow before shipping it, or compare how much the same job costs across several models. For early planning, use conservative token counts. Conversation history, retrieval context, hidden system prompts, and retries often increase the real number of tokens more than expected.
The calculation uses simple public per-million-token rates. It does not apply provider-specific discounts such as cached input, batch API discounts, promotional credits, or enterprise pricing. That simplicity is intentional. For most product planning, a clear baseline estimate is more useful than a complex billing simulator that hides the core economics.