Pangolin releases version 1.22, introducing an AI Gateway that acts as an identity-aware proxy between AI clients and model providers, centralizing access management for both cloud and self-hosted models.

The update strengthens centralized access control by applying the same users, roles, SSO, and access policies across all Pangolin resources, reducing the need to individually expose and secure each remote service.

Gateways can be configured as public or private; public gateways use a fully qualified domain name and Pangolin virtual API keys for authentication, while private gateways are accessible only from devices connected through the Pangolin client and rely on the authenticated client session as the credential.

Community Edition now includes private SSH and HTTPS resources, enabling SSH to internal systems and access to internal web applications via private Pangolin network hostnames.

The interface now displays certificate status information for public resources, allowing users to verify TLS certificate validity directly on resource pages.

Community Edition gains access to previously Enterprise-only features, including browser-based SSH, RDP, and VNC resources, expanding remote access capabilities for all users.

The AI Gateway supports multiple providers—OpenAI, Anthropic, Google Gemini, Amazon Bedrock, Google Vertex AI, Microsoft Foundry, OpenRouter, and Vercel AI Gateway—and self-hosted options like Ollama, vLLM, and Bifrost, with the ability to attach multiple providers to a single gateway.

Additional improvements include the option to send test notifications for alert actions, YAML/container labels for resource definitions, and various bug fixes and interface enhancements.

Readers are directed to Pangolin’s official release announcements and changelog for more details.