A tool that tracks OpenAI API cost per request

The author says they built LLMWatch after receiving an unexpected $200 OpenAI bill. The tool is a proxy that sits between an app and OpenAI. They say changing one baseURL line logs each request with cost, latency, and token breakdown. The free tier includes 1,000 requests per month, while Pro costs $20 per month for unlimited use and cost alerts.

Key points

  • The tool was made after the author got an unexpected OpenAI bill.
  • It connects by changing one baseURL line.
  • It logs cost, latency, and token use for each request.
  • The free tier allows 1,000 requests per month.
  • The Pro plan is $20 per month and adds unlimited use plus cost alerts.

Quick term guide

baseURL
The main web address an app uses when sending requests.
latency
The total time you wait from sending a request to getting a complete response.
free tier
A set amount of usage a service provides at no cost before charging begins.
business
An activity where you provide value to others in exchange for money.
revenue
Revenue is the money a business earns before taking out costs.
prompt
Text instructions you give to an AI tool.
production
The live version of a service that real users use.
privacy
How a tool protects personal data, such as voices and conversation content.
Read original