# Raptor

## Docs

- [How It Works](https://docs.raptordata.dev/features/architecture.md): A Rust-powered proxy designed for speed and reliability
- [Evidence Logging](https://docs.raptordata.dev/features/evidence-logging.md): Full audit trail for every AI request
- [AI Firewall](https://docs.raptordata.dev/features/firewall.md): Block prompt injection before it reaches your model
- [Semantic Cache](https://docs.raptordata.dev/features/semantic-cache.md): Cut AI costs by 60%+ without changing your code
- [Anthropic SDK](https://docs.raptordata.dev/integrations/anthropic.md): Use Raptor with the official Anthropic SDK
- [OpenAI SDK](https://docs.raptordata.dev/integrations/openai.md): Use Raptor with the official OpenAI SDK
- [Vercel AI SDK](https://docs.raptordata.dev/integrations/vercel-ai-sdk.md): Use Raptor with Next.js and the Vercel AI SDK
- [Introduction](https://docs.raptordata.dev/introduction.md): The blazing-fast AI gateway. Built in Rust. ~5ms overhead.
- [Quickstart](https://docs.raptordata.dev/quickstart.md): Start using Raptor in under 2 minutes

## OpenAPI Specs

- [openapi](https://docs.raptordata.dev/api-reference/openapi.json)

## Optional

- [GitHub](https://github.com/raptordata)
- [Discord](https://discord.gg/raptordata)