Auriko

About Auriko

Auriko is an LLM gateway that sits between your application and a dozen model providers — OpenAI, Anthropic, Google AI Studio, xAI, Fireworks, Together, DeepSeek, DeepInfra, MiniMax, Moonshot, Z.AI and SiliconFlow — behind one OpenAI-compatible endpoint. The differentiator it argues for is cost modelling rather than a price table: instead of comparing headline per-token rates, it models how a specific workload interacts with each provider's prompt-caching mechanics and routes each request to whichever provider is actually cheapest once cache behaviour is accounted for. Routing strategies are configurable — optimise for cost, time-to-first-token, throughput or a custom objective, with constraints like a p95 throughput floor, an input-cost ceiling, zero-data-retention providers only, or structured-output-capable models only. It handles automatic failover, load balancing, and key orchestration in either BYOK or platform-key mode, and it exposes budget controls with spend limits and alerts scoped per workspace or per API key. Integration is a base_url swap plus an optional extra_body block for the routing policy, and the documented integrations cover the OpenAI Agents SDK, Claude Agent SDK, Google ADK, LangChain, the Vercel AI SDK, LlamaIndex, CrewAI, Claude Code and several coding agents. The company charges no markup on provider prices at any tier, which puts it in the same billing posture as OpenRouter but with more explicit routing control.

Key Features

Tags