Hi! I'm the creator of **oxideLLM**.

I believe that developers and startups shouldn't be locked into a single AI provider (like OpenAI) or have their apps break during upstream outages. I also believe we shouldn't have to pay massive cloud premiums or run heavy Python/Node proxies just to get basic failover, routing, and usage tracking.

That's why I built **oxideLLM**: a self-hosted, ultra-lightweight LLM gateway written in Rust that adds virtually zero latency (~3% overhead) and handles failover, health checks, and telemetry out of the box.

### Why Support this Project?
oxideLLM is 100% free and open-source under AGPL-3.0. By buying me a coffee, you are helping me cover:
* 🖥️ **Testing Infrastructure**: GPU and API credits to test and validate upstreams (Ollama, vLLM, Groq, Anthropic, OpenAI).
* ⚡ **Development Time**: Keeping the project fast, secure, and shipping features like prompt caching, local Parquet storage, and semantic routing.
* ☕ **Fuel**: Keeping me caffeinated while writing performance-critical Rust code!

If oxideLLM has saved your application from downtime or helped you optimize your AI bills, consider supporting the project. Thank you!