Zyquo Router 1.0.0
The first release of Zyquo Router — your Mac as a local LLM gateway. One
OpenAI-compatible endpoint (http://localhost:8787/v1) fronting 170 models across
12 providers: Anthropic, OpenAI, xAI, Mistral, Google Gemini, Alibaba Qwen, DeepSeek,
Kimi, Perplexity, Together AI, DeepInfra, and Cerebras.
Highlights
- 🌐 Spec-exact OpenAI API — streaming + non-streaming chat completions, tool
calling, vision, and reasoning models; the official OpenAI Python and JS SDKs work
unmodified (verified across all 170 models — seedocs/VERIFICATION.md) - 🔁 Native translation of the Anthropic Messages API and Gemini
generateContent,
plus per-provider quirk normalization for the ten OpenAI-compatible upstreams - 🎛 Control room UI — live dashboard (req/min, tokens, cost, per-provider
breakdown), request inspector with timing waterfall, model aliases, fallback chains,
built-in Playground, in-app API docs, menu bar extra, ⌘K palette - 🔐 Private by design — AES-256-GCM machine-bound key vault (no Keychain),
localhost-only by default, LAN gated behind local API keys, redacted logs - 🍎 100% native — Swift + SwiftUI + SwiftNIO, universal binary (Apple Silicon +
Intel), Developer ID signed, notarized, and stapled
Installation
- Download ZyquoRouter.dmg below
- Open it and drag Zyquo Router to Applications
- Launch → add a provider key in Keys → pick a port → Start
- Point any OpenAI SDK at
http://localhost:8787/v1
Requires macOS 13.0+.