* [OpenWA v0.23.0](https://github.com/rmyndharis/OpenWA) – Self-hosted WhatsApp API gateway with pluggable adapters, multi-session support, dashboard, and webhooks.

* [OmniRoute v3.8.50](https://github.com/diegosouzapw/OmniRoute) – AI gateway providing an OpenAI-compatible endpoint with multi-provider routing, load balancing, retries, rate limits, caching, and observability for cost-aware inference.

* [Bifrost plugins/telemetry/v1...](https://github.com/maximhq/bifrost) – High-performance AI gateway offering multi-provider access with automatic failover, load balancing, and zero-downtime deployments.

* [Conduit v1.17.0](https://github.com/tsouth89/toolport) – Local MCP gateway that centralizes and shares authenticated server tool catalogs across AI clients while reducing per-request token overhead via lazy tool discovery.

* [OpenSurge for Mac v0.2.1](https://github.com/YTwsy/OpenSurge-for-Mac) – Surge-style macOS whole-home gateway and control plane with mihomo TUN, DHCP/DNS, NAT, IPv4 forwarding, and an agent-friendly web GUI and virtual lab.

* [Unkey api/v1.1.27](https://github.com/unkeyed/unkey) – API authentication and authorization platform for modern developer workflows.

* [Ferro Labs AI Gateway v1.4.5](https://github.com/ferro-labs/ai-gateway) – High-performance Go-based AI gateway routing OpenAI-compatible LLM requests across 30+ providers with caching, guardrails, and cost controls.

* [Tingly Box v0.260827.0-rc1](https://github.com/tingly-dev/tingly-box) – High-performance desktop LLM proxy that unifies access to many models and providers via a single OpenAI-compatible API.