Back

hii actually I have my ai application. and I used gemini for the ai model. as am scaling my application I had to use gemini more. so what I did was I created multiple projects and in each project i created a single key such that the limits are only

I would suggest using proxy for AI api request however proxy has it's own problems and I created a package on npm called llmpool it can loadbalance multiple LLMs so you can use this for each of your feature and for every feature use groq openai whate