Current settings:
Requests per second · RPM
Concurrent requests / sec: 15
RPM: 600
But @alchigamid can actually keep getting 429s
amazing! Is my limit too low? I don’t understand.
I’d like to hear everyone’s opinions, as well as @alchigamid’s.
I’m quite curious how everyone uses it
I’m using oh-my-pi to refactor mathematical code. I can’t install RPM packages in it. So I hope this is a server limitation intended to prevent the server from being overloaded.
An average of 4,300 requests per day, 1.2 billion prompt tokens, 160,000 tokens per request on average (up to 890,000), almost exclusively using muse-spark-1.3, running nonstop for 24 hours, with 7530/7532 being streaming. This doesn’t really look like “math code,” does it?
Is the guy doing RL training for the formal verification of mathematics?
I am developing a plugin for oh-my-py that searches for engineering-related articles. It can collect and analyze articles, help you write mathematical code in Julia, and verify the correctness of the code based on data from the articles. I hope this plugin can contribute to the scientific community. https://github.com/prestalab/science-engineering-research-omp
muse-spark-1.3 handles mathematical problems very well. The mathematical code generated by muse-spark-1.3 is stored in a private repository, so I cannot display it.
I set an RPM limit in the API key settings, but it was not saved.
I’m sorry, but I couldn’t find the repository on your GitHub.
And according to your account’s commit history,
your most recent commit was on September 21, 2026.
I’m unable to verify this here….
Please provide proof by other means that you did indeed spend time working on this project—even a screenshot of you using oh-my-pi to complete the project would suffice (you may redact information if necessary, including your oh-my-pi usage statistics).
If that provides sufficient proof,
I’ll fully support your work!