usestrix
[BUG] LLM Request Failed - 429 rate limit
I tried using the Gemini 2.5 Pro and Gemini 2.5 Flash models but got rate limited so early. Is there any feature to use multiple API keys at once?
Hi @xalgord 👋,
I’d love to help with this issue!
From the error logs, it looks like the LLM request is hitting a 429 rate-limit when using the Gemini 2.5 / 2.5 Flash models.
This usually happens because the current API key quota is exhausted or requests are being sent too quickly without proper throttling.
🧩 Possible Fix Ideas:
- Implement a retry/backoff mechanism for 429 responses.
- Add support for rotating multiple API keys (as you mentioned) — for example, cycling through keys stored in
.envvariables. - Add a configurable request delay or queue system in Strix to avoid hitting provider limits.
- Log the number of requests sent per minute to help monitor usage.
I don’t have my laptop right now, but I’d be happy to work on this fix collaboratively once I’m back on my setup.
Please assign or tag me when it’s open for contributors 🙌
Thanks!
— @Armankb2
also which model would be better for web application testing, chatgpt 5 or gemini 2.5 pro?
hello i tried to use the gemini flash 2.5 but i have this problem : export STRIX_LLM="google/gemini-2.5-flash"
@fortinixeyas it should be export STRIX_LLM="gemini/gemini-2.5-flash"
You can see other models at https://docs.litellm.ai/docs/
hello i tried to use the gemini flash 2.5 but i have this problem : export STRIX_LLM="google/gemini-2.5-flash"
You can change it to export STRIX_LLM="gemini/gemini-2.5-flash" and the problem should be fixed.
hello i tried to use the gemini flash 2.5 but i have this problem : export STRIX_LLM="google/gemini-2.5-flash"
You can change it to export STRIX_LLM="gemini/gemini-2.5-flash" and the problem should be fixed.
i got this issue , can u help me
