Report: Low operating speed

[AnsirU]

@WeihaoTan @DVampire My PC uses Nvidia 3060ti GPU, and I called Gpt-4o model via the azure openai api.
However, when the agent executes the task of using chrome, it takes several minutes to decide the next step.
I wonder if it is the right operating speed? Or there is something wrong with my configuration?

[WeihaoTan]

Sorry for the late response. Cradle does not need to rely on the local GPU. It mainly depends on the latency between you and OpenAI server. You may go to the OpenAI API website (https://platform.openai.com/docs/overview) to run a minimal example to call GPT-4o. If it takes a long time (e.g., longer than 30s), you might need to check your connection issue and try to change your VPN node.