epam/ai-dial-core

The main component of AI DIAL, which provides unified API to different chat completion and embedding models, assistants, and applications

What it solves

DIAL Core acts as a unified HTTP proxy that simplifies interaction with various chat completion and embedding models. It eliminates the need for applications to integrate with multiple different AI model APIs individually by providing a single, standardized API surface.

How it works

Built with Java 21 and Eclipse Vert.x, the system functions as a gateway. It exposes a set of REST endpoints (defined via an OpenAPI specification) that route requests to different AI models and applications. It includes built-in support for identity providers for authentication and access control, and integrates with MCP servers for toolset execution.

Who it’s for

Developers and platform engineers who need to manage and proxy requests to multiple AI models and embedding services within an enterprise environment, requiring centralized authentication and analytics.

Highlights

  • Unified API: Provides a single entry point for different chat and embedding models.
  • Enterpise-ready Auth: Supports multiple identity providers with configurable role-based access control.
  • Observability: Includes detailed analytics logging for requests, including configurable claim collection and header filtering.
  • High Performance: Leverages Java 21 virtual threads for blocking tasks and the reactive Vert.x framework.
  • Kubernetes Native: Provides Helm charts for streamlined deployment on Kubernetes clusters.

Related

  • Project
    BenedictKing/ccx

    A high-performance AI API proxy and gateway that unifies Claude, OpenAI, and Gemini APIs into a single entry point with built-in channel management and failover.

  • Project
    DEEIX-AI/DEEIX-Chat

    An open-source AI platform that provides a unified interface for model routing, multimodal chat, and enterprise operations like billing and identity management.

  • Project
    withoneai/cli

    A CLI tool that gives AI agents authenticated access to 750+ platforms via a single interface, eliminating the need to manage multiple API keys and OAuth flows.

  • Project
    mock-server/mockserver-monorepo

    An HTTP(S) mock server and proxy for testing that simulates API dependencies, records traffic, and provides specialized mocking for AI/LLM chat-completion APIs.

  • Project
    danny-avila/LibreChat

    A self-hosted AI chat platform that unifies multiple AI providers into a single, privacy-focused interface with support for agents, code execution, and multi-user management.