I’ve been spending a lot of time exploring AI tools and working with prompts recently.
One thing I noticed is that it’s not always easy to understand how much a prompt actually costs when using AI APIs.
You write a prompt, send it to a model, get a response, and move on.
But what if you’re testing multiple prompts or comparing different AI models? Keeping track of token usage and API costs can become confusing.
Why I Built It
I wanted a simple place where I could:
- Count the tokens in a prompt.
- Estimate API costs before sending requests.
- Compare the costs of different AI models.
- Make prompts shorter and more efficient.
Instead of constantly switching between different tools, I decided to build a small project that brings these features together.
Introducing PromptBurn
PromptBurn is a free tool designed for people who work with AI prompts and APIs.
It currently includes:
- Token Counter
Check how many tokens your prompt contains.
- API Cost Calculator
Estimate potential API costs based on token usage and model pricing.
- Model Comparison
Compare costs across supported AI models.
- Prompt Compression
Experiment with making prompts shorter while keeping their meaning.
Privacy Was Important to Me
I also wanted to keep the experience simple and privacy-conscious.
PromptBurn is designed to process prompt text directly in the browser, rather than sending it to a server.
What’s Next?
This is still a work in progress.
I’d like to improve the accuracy, add useful features, and make the experience better for developers working with AI APIs.
You can check it out here:
I’d genuinely appreciate feedback from other developers.
A few questions:
- What would you improve first?
- What feature is missing?
- How do you currently track token usage and API costs?
Thanks for reading!
