Awesome GitHub RepositoriesAPI Gateways

This project is a self-hosted large language model chat interface and AI model aggregator. It provides a unified web environment for interacting with multiple AI providers and local models, acting as a provider-agnostic API gateway to standardize requests across different endpoints. The platform functions as an agentic AI framework and generative UI workspace, enabling the construction of specialized assistants with custom instructions and subagents. It features a sandboxed code interpreter for secure execution of multiple programming languages and a generative UI system that renders interact

This project is an AI-focused API gateway and proxy system designed to intercept, standardize, and route requests across heterogeneous language model providers. It functions as a middleware layer that normalizes incoming traffic and manages authentication, ensuring consistent integration across diverse service interfaces. The system features a programmable routing engine that executes user-defined scripts to evaluate request content in real-time. This allows for dynamic traffic management, where requests are inspected, transformed, and redirected to specific model endpoints based on custom lo

Medusa is a headless commerce engine designed as a modular, API-first platform for building custom digital storefronts and business applications. Its architecture is built on a decoupled system where core business logic is encapsulated into independent, swappable modules that communicate through defined interfaces, allowing developers to incrementally adopt or replace components to fit specific operational needs. The platform distinguishes itself through a highly extensible design that supports complex commerce requirements, including multi-vendor marketplace operations, B2B purchasing workfl

Agent-Reach is an AI agent web gateway and search tool that provides language models with the ability to search and read content from the open web, social media, and community forums without using official APIs. It functions as a routing layer that connects large language models to various internet backends while managing content parsing and connection health. The system enables API-free information retrieval by using open-source backends to extract text and metadata from platforms such as Twitter, Reddit, and YouTube. It converts unstructured website content, RSS feeds, and video transcripts

This project is a self-hosted dashboard portal designed to centralize access to internal applications and infrastructure services. It functions as a configuration-driven platform that automatically discovers and organizes services from container runtimes and cluster management systems, presenting them within a unified, customizable web interface. The system distinguishes itself through a declarative widget framework that allows users to construct dashboard components by mapping raw API responses to visual elements. It includes a secure internal proxy layer that handles authentication, header

Budibase is a low-code application platform and enterprise internal tool builder used to create custom business applications for organizational processes and reporting. It functions as a self-hosted backend as a service, providing the infrastructure to manage database integrations and expose public data interfaces for external application access. The platform includes an AI agent orchestrator for deploying autonomous agents that interact with business data and execute operational tasks. It differentiates itself through self-hosted infrastructure management, allowing the system to run on priva

Zheng is a Spring Boot microservices framework and enterprise J2EE development platform. It functions as a distributed service gateway and identity provider, providing a foundation for building complex business applications and microservices infrastructure. The project includes a comprehensive enterprise content management system and an OAuth2 identity provider for managing single sign-on and third-party social login integrations. It also features a MyBatis ORM code generator that automatically creates database models and boilerplate functions from existing tables. The platform covers a broa

cc-connect is an AI agent messaging bridge and session manager that connects local AI coding agents to third-party messaging platforms. It acts as a multimodal AI chat relay and a OneBot protocol gateway, allowing users to control local AI agents remotely via a variety of chat interfaces. The project distinguishes itself by providing a remote AI agent controller that enables the management of agents through slash commands and a web management dashboard. It supports multi-tenant project orchestration and session-based context isolation, ensuring that independent conversation threads are mainta

mi-gpt is a voice assistant bridge and agent orchestrator that connects smart speakers to large language models. It functions as an integration layer that routes audio requests from hardware speakers to AI providers and converts generated text back into speech via a customizable synthesis system. The project features a retrieval-augmented generation knowledge base that uses embeddings and external documents to provide context-aware responses. It includes a persona definition system for configuring behavioral rules, system prompts, and roleplay characteristics, alongside a plugin architecture

This project is a comprehensive e-commerce platform implementation available as a Spring Boot application, a Spring Cloud microservices architecture, and a version rewritten in the Go programming language. It provides a full-stack retail system featuring a Vue 3 storefront interface and a centralized backend administration portal. The platform is specifically designed to handle high-concurrency flash sales and coupon distribution systems to manage sudden spikes in purchase requests. It supports multiple deployment strategies, ranging from monolithic server-side rendering to a decoupled fronte

This project is a container image registry and server-side storage system designed to house container images, layers, and manifests. It functions as an OCI compliant registry server that adheres to the Open Container Initiative Distribution Specification to store and deliver content over HTTP. The system provides a self-hosted solution for managing private libraries of container images within professional-grade infrastructure. It is designed to enable the development of custom registries by extending a base toolkit with specialized libraries and business logic. The registry covers image dist

Spinnaker is a multi-cloud continuous delivery platform designed to automate software releases and deployment pipelines across various public cloud providers and Kubernetes clusters. It functions as a cloud deployment orchestrator and infrastructure delivery tool, coordinating the promotion of software artifacts through multiple environments using visual workflows and directed acyclic graphs. The platform distinguishes itself with a dedicated canary analysis engine that compares performance metrics between new and stable software versions to automate release decisions. It utilizes cloud-agnos

CoAI is an enterprise-grade, self-hostable AI gateway platform that unifies access to over 200 AI models from more than 35 providers through a single OpenAI-compatible API endpoint. It functions as a multi-tenant gateway, routing requests across providers with load balancing, automatic failover, and priority-based routing, while exposing standard OpenAI API endpoints for chat, image generation, model listing, and billing to enable seamless integration with existing tools and clients. The platform distinguishes itself through a comprehensive set of operational capabilities built around the gat

Shenyu is a microservices API gateway designed to route external traffic to backend services using dynamic rules and protocol conversion. It functions as a central entry point that manages traffic flow through a combination of an API traffic governor, a distributed configuration manager, and a security layer for protecting endpoints. The project features a dynamic plugin architecture that allows for the injection of custom request processing logic without restarting the server. It utilizes a distributed coordination service to synchronize routing and policy updates across a gateway cluster in

Higress es una API gateway nativa de IA y nativa de la nube que enruta, asegura y optimiza el tráfico entre clientes y servicios de grandes modelos de lenguaje. Funciona como un punto de entrada centralizado para microservicios, sirviendo tanto como controlador de ingreso (ingress) de Kubernetes como orquestador de puerta de enlace de IA. El proyecto se distingue por gestionar el tráfico a través de múltiples proveedores de IA utilizando un protocolo unificado, incorporando limitación de tasa consciente de tokens y almacenamiento en caché de respuestas para optimizar la inferencia del modelo. Coordina la comunicación entre modelos de IA y herramientas externas para proporcionar contexto y datos en tiempo real, al tiempo que aloja puntos finales de servidor para agentes de IA. Sus capacidades incluyen la aplicación de seguridad de API mediante firewalls de aplicaciones web, gestión automatizada de certificados TLS y descubrimiento dinámico de servicios. La puerta de enlace admite el procesamiento de solicitudes personalizadas a través de plugins de WebAssembly en sandbox que permiten la transformación del tráfico con recarga en caliente. El sistema implementa API de ingreso estandarizadas para gestionar el enrutamiento de red dentro de clústeres en contenedores con baja sobrecarga de recursos.

Llama-stack es un stack de orquestación estandarizado y una puerta de enlace de API para IA generativa. Proporciona una capa de comunicación unificada y una interfaz consistente para desplegar, gestionar e interactuar con varios proveedores y despliegues de modelos de lenguaje de gran tamaño. El sistema funciona como un framework de agentes que gestiona la ejecución de herramientas y paquetes de habilidades versionados para automatizar tareas complejas. Incluye un sistema de procesamiento por lotes para manejar grandes volúmenes de solicitudes asíncronas mediante procesamiento offline y una interfaz de base de datos vectorial para almacenar y buscar documentos, permitiendo la generación aumentada por recuperación (RAG). El stack cubre capacidades de alto nivel, incluyendo la orquestación de agentes de IA, el despliegue de modelos y la estandarización de APIs de modelos para permitir el cambio entre proveedores sin reescribir el código de la aplicación.

BentoML is a machine learning model serving framework and GPU-accelerated inference server designed to package, deploy, and scale AI models as production-ready REST APIs. It functions as an AI model lifecycle manager and an inference graph orchestrator, enabling the chaining of multiple models and custom logic into complex pipelines for advanced task sequences. The framework distinguishes itself through a dynamic batching engine that optimizes GPU throughput and an artifact-based packaging system that bundles model weights and dependencies into immutable archives for consistent deployment. It

BAML is a prompt engineering framework and LLM client generator that defines AI prompts as type-safe functions. It serves as a structured data extraction tool and workflow orchestrator, transforming unstructured model responses into strongly typed objects using a custom schema language and alignment algorithms. The project distinguishes itself by using a compiler to generate language-specific boilerplate code for API communication and output parsing. It features a dedicated environment for designing complex prompt templates with conditional logic and reusable snippets, and employs genetic alg

SpringCloud-Learning is an educational project that demonstrates how to build microservices using Spring Cloud, covering the core patterns of service discovery, centralized configuration management, and API gateway routing. The project provides hands-on examples for registering and discovering microservice instances with Nacos, Eureka, or Consul, and for routing external API requests through Spring Cloud Gateway with support for filters and load balancing. The tutorials explore service resilience through circuit breakers and rate limiting with Sentinel and Hystrix, including custom fallback l

Aidea es un cliente de IA multiplataforma y un proyecto de infraestructura autohospedada. Consiste en una aplicación Flutter y un sistema backend contenedorizado diseñado para proporcionar una interfaz unificada para interactuar con modelos de lenguaje de gran tamaño y servicios de generación de imágenes por IA. El proyecto funciona como una puerta de enlace de IA autohospedada, permitiendo a los usuarios gestionar y desplegar instancias privadas de modelos de lenguaje e imagen en su propio hardware. Esta arquitectura permite el enrutamiento de diversas consultas de datos a través de una API gateway estandarizada a múltiples proveedores de IA mientras se mantiene el control sobre el almacenamiento de datos y la configuración del servidor. El sistema cubre flujos de trabajo conversacionales, difusión de texto a imagen y procesamiento de flujos asíncronos. Utiliza un único código base para renderizar una interfaz de usuario consistente en entornos móviles, de escritorio y web, respaldado por la orquestación de servicios basada en contenedores para el backend.