Convert Raycast AI API to OpenAI-compatible API. Built with HonoJS, supports Raycast V2 authentication, deployable to Cloudflare Workers / Docker / Local Binary.

This project is based on raycast-relay, which has been archived and is no longer maintained. The original project does not support the new Raycast V2 authentication system that is now required. This version utilizes Hono and provides support for Raycast V2 authentication.

⚠️ Be aware that not setting the API_KEY environment variable leaves your API open to anyone.

docker run -d \
  --name raycast2api \
  -p 3000:3000 \
  -e RAYCAST_BEARER_TOKEN=your-raycast-bearer-token \
  -e RAYCAST_DEVICE_ID=your-raycast-device-id \
  ghcr.io/xxxbrian/raycast2api:latest

apiVersion: apps/v1
kind: Deployment
metadata:
  name: raycast2api
spec:
  replicas: 3
  selector:
    matchLabels:
      app: raycast2api
  template:
    metadata:
      labels:
        app: raycast2api
    spec:
      containers:
      - name: raycast2api
        image: ghcr.io/xxxbrian/raycast2api:latest
        ports:
        - containerPort: 3000
        env:
        - name: RAYCAST_BEARER_TOKEN
          valueFrom:
            secretKeyRef:
              name: raycast-secret
              key: bearer-token
        - name: RAYCAST_DEVICE_ID
          valueFrom:
            secretKeyRef:
              name: raycast-secret
              key: device-id
        - name: API_KEY
          valueFrom:
            secretKeyRef:
              name: raycast-secret
              key: api-key
        livenessProbe:
          httpGet:
            path: /health
            port: 3000
          initialDelaySeconds: 10
          periodSeconds: 30
        readinessProbe:
          httpGet:
            path: /ready
            port: 3000
          initialDelaySeconds: 5
          periodSeconds: 30

Download the latest release from Releases and run it.

Use proxy tools (like Proxyman, Charles) to capture Raycast app network requests:

1. Set up proxy and trust certificate
2. Add backend.raycast.com to SSL proxy list
3. Send AI request in Raycast
4. Extract from request headers:
   • Authorization: Bearer <token> → RAYCAST_BEARER_TOKEN
   • X-Raycast-DeviceId → RAYCAST_DEVICE_ID

Deployed API URL: https://your-worker.your-account.workers.dev/

• GET /health - Health check (liveness probe, no auth required)
• GET /ready - Readiness check (readiness probe, no auth required)
• GET /v1/models - Get model list (requires auth if API_KEY is set)
• POST /v1/chat/completions - Chat completion (OpenAI compatible, requires auth if API_KEY is set)

Implements Raycast V2 signature authentication:

• ROT13+ROT5 encoding
• HMAC-SHA256 signing
• Timestamp validation
• Device ID binding

src/
├── index.ts           # Main entry
├── config.ts          # Configuration
├── types.ts           # Type definitions
├── utils.ts           # Utils and auth
└── handlers/
    ├── chat.ts        # Chat handler
    └── models.ts      # Models handler
server.ts              # Bun server entry for local binary

bun run dev        # Cloudflare Workers dev mode
bun run dev:local  # Local dev mode

• raycast-relay
• raycast2api