Dynamic Routing
Let CommandLyne pick the best model per task on quality, speed and cost, covering routing strategies, configuration and how routing decides.
Overview
Dynamic Routing allows CommandLyne to automatically choose the best AI model for each task based on quality, speed, and cost.
Core Configuration
-
1
Enable Dynamic Routing
This is enabled by default in the guide.
-
2
Auto-upgrade on negative feedback
Allows the agent to switch to a stronger model if the response is unsatisfactory.
Routing Strategies
-
1
Balanced
Mixes quality, speed, and cost.
-
2
Cost-first
Uses the lowest-cost model that meets basic quality.
-
3
Quality-first
Uses a more advanced model for better output.
How Routing Works
-
- Complex tasks use higher-intelligence models.
- Simple tasks use lower-cost models.
- If a model fails, CommandLyne can reroute requests to another available model.
Save Changes
Scroll down and click Save after making changes.
