Auriko

Auriko is an AI inference control plane for teams running production LLM workloads. It provides a unified OpenAI-compatible API across model providers, then routes calls using cost, latency, throughput, provider health, cache behavior, zero-data-retention constraints, and structured-output requirements. The official site positions it as an LLM gateway with routing, observability, FinOps, budget controls, failover, key orchestration, and global deployment for OpenAI, Anthropic, Google AI Studio, xAI, Fireworks, Together, DeepSeek, DeepInfra, MiniMax, Moonshot, Z.AI, and SiliconFlow.

• Unified model API — access multiple providers through one OpenAI-compatible endpoint • Cost-aware routing — optimize requests based on workload, provider pricing, and prompt caching mechanics • Reliability controls — automatic failover, provider health signals, and capacity intelligence • FinOps features — workspace/API-key spend limits, alerts, and routing objectives • Production deployment — global edge routing, BYOK/platform keys, and policy constraints such as ZDR providers