Decision tools and launches with source-aware context.

Filter source-backed model routes by workload, context, capabilities, and budget.

Estimate monthly input, output, cache-read, and per-request charges.

Estimate model-weight and optional architecture-aware KV-cache memory.

Every tracked model route with input, output and cache prices, plus observed price changes.

Compare context windows, output limits and modalities across tracked model routes.

Put your own GPU rate against API token prices for the same monthly workload.