LLM Gateway

LLM Gateway sits between an application and large language model providers such as OpenAI, Anthropic and Google. Developers keep their existing OpenAI SDK code, change the base URL, and can then call more than 250 models from over 40 providers with a single key, switching models without code changes. The service handles automatic failover, caching, spend caps and scoped keys, and adds guardrails against prompt injection and personal data leaks. Dashboards break down requests, tokens, errors, latency and cost by project, model and provider, and enterprise plans add audit logs and data retention controls. Teams can use the hosted cloud, bring their own provider keys, or self-host the whole stack with Docker.