Switchyard
An AI model routing middleware that acts as an OpenAI API-compatible proxy layer. It dynamically selects the optimal option among multiple model APIs—both cloud-based and local—based on response quality, inference speed, and call cost, supporting load balancing across local and cloud environments.
Timeline (last 90 days)
NVIDIA's Switchyard is a model switching/routing tool that can be integrated into Claude Code.
Switchyard adopts an escalation routing strategy: requests are first sent to the cheapest model, and if the response quality is insufficient, they are escalated step by step to more expensive models such as Haiku, Sonnet, and Opus.
PAIR handles local inference traffic scheduling for hardware entirely under the user's control, while Switchyard selects models at the API level based on quality and cost and handles decisions that may leave the local network.
All Facts (3)
NVIDIA's Switchyard is a model switching/routing tool that can be integrated into Claude Code.
50%UnverifiedSwitchyard adopts an escalation routing strategy: requests are first sent to the cheapest model, and if the response quality is insufficient, they are escalated step by step to more expensive models such as Haiku, Sonnet, and Opus.
50%UnverifiedPAIR handles local inference traffic scheduling for hardware entirely under the user's control, while Switchyard selects models at the API level based on quality and cost and handles decisions that may leave the local network.
50%