digiRunner is an enterprise-grade API Gateway designed to unify microservices and bridge infrastructure to AI services, specifically addressing the challenges of integrating Generative AI. It targets developers and operations teams seeking to manage AI models securely, control costs, and avoid vendor lock-in, offering a governed control plane for both traditional APIs and LLM orchestration.

How It Works

digiRunner functions as an AI Gateway, abstracting diverse AI providers (e.g., OpenAI, Azure, Ollama) through a universal interface. This allows applications to interact with a single digiRunner endpoint, enabling seamless switching of backend models without code redeployment. It enhances security by managing provider API keys centrally and implements FinOps and Tokenomics control, moving beyond simple rate limiting to enforce granular input/output token limits and flexible cost policies. Additionally, it treats prompt engineering as infrastructure, offering a template registry for versioned, managed prompts.

Quick Start & Requirements

• Installation: Options include Docker (docker run -it -d -p 31080:18080 tpisoftwareopensource/digirunner-open-source), Docker-Compose, Kubernetes, Helm, or a standalone executable downloaded from releases. Building from source requires OpenJDK 25.
• Prerequisites: Minimum 2 CPU Cores and 4 GiB RAM. OpenJDK 25 for building.
• Access: Navigate to http://localhost:31080/dgrv4/login or http://localhost:18080/dgrv4/login with default credentials: username: manager, password: manager123.
• Links: Official documentation, blog, and community resources are available via links in the README.

Highlighted Details

• Zero Vendor Lock-in: Universal interface for AI providers allows easy backend model switching.
• FinOps & Tokenomics Control: Granular cost management for LLMs via token limits and flexible policies.
• Prompt Engineering as Infrastructure: Centralized, versioned prompt management through a template registry.
• Core API Management: Supports RBAC, API keys, OAuth2/OIDC, intelligent routing, and traffic shaping.

Maintenance & Community

The project fosters community engagement, with a badge system recognizing contributors and links to community resources and LinkedIn.

Licensing & Compatibility

The specific open-source license is not explicitly stated in the provided README, which is a critical omission for adoption decisions. Compatibility for commercial use or closed-source linking is not detailed.

Limitations & Caveats

The standalone Quickstart executable may trigger security warnings on macOS and Windows due to a lack of official developer program registration. The absence of a clearly defined license is a significant adoption blocker requiring immediate clarification.