Each AgentProfile in AI Hub points to exactly one ModelConfig. You can create multiple ModelConfigs pointing to different providers and models. Different agents in the same system can use different models.
Your Challenge
Your team has access to two models: a powerful cloud API (accurate, costs money per call) and a local Ollama model (fast, free, less capable). You need to serve two agents:
A) A classifier that processes 500 support emails/day (simple categorization task)
B) A contract analysis agent used 10 times/month (needs precise reasoning)
Which model would you assign to each agent in AI Hub? Explain your reasoning, including the cost/quality tradeoff.