5 ms·
Show HN: LLM.Report – A free OpenAI API analytics dashboard
Hi!
My OpenAI costs started going up recently and I hated how I couldn't figure out which model was costing me more?
The OpenAI usage page (https://platform.openai.com/account/usage https://platform.openai.com/account/usage) gives a nice overall breakdown, but it doesn't go into detail about cost / model, token usage, and # requests.
So I reverse engineered their API to make a better dashboard, and then made it free to use.
Try it out and let me know what you think. Also please tell me what other features you wish existed here https://platform.openai.com/account/usage https://platform.openai.com/account/usage but doesn't yet.
- novoreorx 3y agoI am absolutely loving this! I'm currently maintaining an AI chatting app for a small group of people, so I pay a lot of attention to my OpenAI usage. Before, I used Waige[^1] which only gave me a glimpse of how much was used in the current month, but I had no idea about the details of the models or tokens under the hood. Now, with LLM.Report, I can finally see what the users are doing. Here I'd like to share the breakdown of my usage from the last 30 days: - The total cost was $105, with $93 used by GPT-4 - Requested 5,871 times for GPT-3.5 and 1,644 times for GPT-4 - Tokens used 5,880,761 for GPT-3.5 and 2,469,007 for GPT-4 It's interesting to note that even though only 1/3 of the requests are from GPT-4, it accounted for almost 90% of the cost. With the charts available on LLM.Report, I'll have a better insight into how to promote the product and determine new features. This app is a great value to any OpenAI product maker. I'm willing to pay for this app as long as it can provide more analytics on the historical data. [1]: https://pyrolyse.gumroad.com/l/waige https://pyrolyse.gumroad.com/l/waige