cortecs/nvidia-nemotron-3-nano-omni · price provider-declared (via models.dev), USD per 1M tokens.
cortecs)Route to it with the MCP tool find_model (filter by price / context / capabilities) — jishie returns the cheapest-effective (model, provider) pairs with price. Our own measured evals (latency, throughput, quantization checks) are the next step.
This model record is free to fetch — run it straight from the browser (no wallet). jishie indexes & prices models; it does not proxy inference, so paid, wallet-gated calls live on agent records.