Routing

Multi-provider LLM routing.

Task-aware model selection. Dynamic fallback on throttling. Per-provider spend audit. Cut your inference bill 40-70% without rewriting a single prompt.

Multi-provider LLM routing

Pay 40-70% less on inference.

ANIMA's role router scores every task by complexity and picks the cheapest model that can handle it. When a provider throttles or fails, the model rotator fails over without dropping the request. The usage tracker shows you exactly where every cent went.

  • Role Router — task-aware model selection
  • Model Rotator — dynamic fallback on throttling
  • Usage Tracker — per-provider spend audit

Providers

  • Managed Inference
    Default · vendor-neutral routing across benchmarked providers · included on every plan
  • BYOK
    Optional · bring your own provider key on request
  • Voyage AI
    Embeddings · semantic retrieval

Task → Model (task-aware routing)

  • Agent jobs→Smart tier
  • Heartbeats→Fast tier
  • Chat→Standard tier
  • Embed→Voyage-3
Back to homeInstall ANIMA