AI model optimizer: the right model for each use case
The Model Optimizer chooses the most appropriate model for each business use case: lower-cost models handle everyday work in conjunction with your enterprise data doing the heavy lifting, and frontier models are limited to just the users and use cases that need them. Set a default model for each agent and decide which users may switch to premium ones.
AI Cost Intelligence
Model Optimizer
Set a default model for each agent and decide which users may switch to premium ones. Most everyday work runs well on lower-cost models when it draws on your own data.
Questions you can ask
Ask the Model Optimizer agent in plain language; every answer shows its sources.
- Which agents are running on frontier models today?
- Can the policy Q&A agent run on a lower-cost model?
- Who is allowed to switch to a premium model?
- Which agents could move to a lower-cost model without losing quality?
- Set the default model for the policy Q&A agent.
- Which use cases need a frontier model, and which don't?
Example output
It shows the default model set for each agent, such as the fast, low-cost model here, with premium models limited to the users and use cases that need them.
Model per agent
- Fast, low-cost model
- Balanced model
- Most capable model
Illustrative example with sample data.
How it works
- 1
Powered by Flexible LLM Choice
Choose the most appropriate model for each business use case.
- 2
Set a default per agent
Set a default model for each agent.
- 3
Let your data do the heavy lifting
Most everyday work runs well on lower-cost models when it draws on your own data.
- 4
Limit frontier models
Decide which users may switch to premium models, so frontier models go only to the users and use cases that need them.
More AI Cost Intelligence agents
- Agent Usage and Cost AnalyzerSee which users, teams and use cases drive AI costs, and drill down to individual conversations to find what's really driving token spend.
- Agent ROI CalculatorUse cost and usage data to find the agents with real business impact, and phase out the ones without it.
- LLM IndependenceChoose from leading AI models with no vendor lock-in, and switch whenever business needs change or a better-value model launches. Leverage lower-cost open-weight models.
See the Model Optimizer agent at work
Questions about AI Cost Intelligence or this agent? Send us a note and we'll get back to you by email.