Jev Router is a routing layer that selects the appropriate model and reasoning effort for each request from TypeSafe's Jev model family.
- Strengths
- Automatically balances output quality, latency, and resource use by matching request complexity to the right inference profile.
- Best for
- Workloads with variable complexity where you want adaptive inference without manually specifying effort levels or routing logic.
- Limitations
- Adds a routing overhead that may introduce small latency increases, and the selection heuristic may not match your specific quality or speed preferences.
Pricing
Priced per 1M tokens — not yet tracked
This model bills per 1M tokens. That price isn't in the catalog yet.