Hello everyone,
I am completely new to the AI world and I am currently trying to get my first AI agent up and running at home on my Linux machine. I am using the open-source project OpenClaw for this.
To test things out, I set up a free Google Cloud account with the $300 starter credit and generated an API key for Gemini there. I am located in Europe.
My problem:
As soon as I start the agent in the terminal and send my first message, the connection drops and I repeatedly get Error 429 (Too Many Requests / Rate Limit Exceeded).
Looking at the behavior, it seems the OpenClaw agent sends a pretty large chunk of background data/context (around 45,000 tokens) right from the start.
My questions to the community:
Is this generally due to extremely low limits (TPM / Requests per minute) within the Google Cloud Free Trial tier, or did I configure something incorrectly?
Is there any way to increase these limits within the $300 credit without immediately risking actual charges on my credit card?
Has anyone here successfully managed to run OpenClaw using the Google Free Tier?
Thank you so much in advance for helping out a Linux and AI beginner!
Best regards
D.D.