What is Backmesh?
Backmesh is an open-source backend solution designed to protect your LLM API keys from exposure in client-side applications. By acting as a secure proxy, it ensures that secret keys are never shipped to users' devices, preventing costly leaks that can result in thousands of dollars in unauthorized API usage. Backmesh integrates with your existing authentication provider, verifying requests via JWTs so that only authenticated users can access the LLM API through your app. This eliminates the risk of API key theft from decompiled code or network inspection.
Key features include configurable per-user rate limits, allowing you to set restrictions such as a maximum of five OpenAI API calls per user per hour. This granular control prevents abuse and helps manage costs effectively. Additionally, Backmesh protects sensitive API resources like Files and Threads by enforcing ownership—only the user who created a resource can access it, ensuring data privacy and security.
Backmesh uses military-grade encryption to safeguard your LLM API key in transit and at rest. The API Gatekeeper intercepts all calls from your app, validates the user's JWT, applies rate limits, and forwards the request to the LLM provider. All interactions are instrumented, providing detailed usage logs that help you identify patterns, reduce costs, and improve user satisfaction. Whether you're building a chatbot, content generator, or any AI-powered feature, Backmesh offers a robust, easy-to-deploy security layer. It is thoroughly tested and can be self-hosted or used as a managed service, giving you flexibility and control over your AI infrastructure.
Key Features
LLM API key protection secure proxy JWT authentication per user rate limits resource ownership enforcement military grade encryption cost abuse prevention client side securityWho is it for?
frontend developers, backend engineers, security teams, SaaS startups, AI application builders, platform engineers
