Routing Modes
Overview
Section titled “Overview”TokenRouter offers four routing modes to optimize for different priorities.
Mode Comparison
Section titled “Mode Comparison”| Mode | Use Case | Cost Priority | Quality Priority | Latency Priority |
|---|---|---|---|---|
auto:balance | General purpose | 40% | 40% | 20% |
auto:cost | High volume | 80% | 10% | 10% |
auto:quality | Complex tasks | 10% | 80% | 10% |
auto:latency | Real-time | 10% | 10% | 80% |
When to Use Each Mode
Section titled “When to Use Each Mode”Balance
Section titled “Balance”- Chat applications
- General Q&A
- Content generation
- Batch processing
- Simple classification
- High-volume operations
Quality
Section titled “Quality”- Code generation
- Research
- Complex reasoning
Latency
Section titled “Latency”- Real-time chat
- Interactive apps
- Quick lookups