chatgpt-web-dev/chatgpt-web
Chatgpt Web
Este proyecto es una interfaz web autohospedada diseñada para interactuar con modelos de lenguaje de gran tamaño. Proporciona un panel centralizado que permite a los usuarios gestionar las comunicaciones con los modelos, mantener historiales de conversación persistentes y organizar bibliotecas de prompts dentro de un entorno privado y contenedorizado.
La plataforma se distingue por sus sólidos controles administrativos y de seguridad, incluyendo soporte para proveedores de identidad empresariales, acceso basado en contraseñas y limitación de tasa de solicitudes. Facilita una conectividad flexible al permitir a los usuarios configurar proxies de red personalizados, gestionar múltiples claves API y ajustar la configuración del razonamiento del modelo y la ventana de contexto para adaptarse a requisitos operativos específicos.
La aplicación admite un conjunto completo de funciones para gestionar interacciones con IA, como respuestas en streaming en tiempo real, renderizado de contenido enriquecido para código y notación matemática, y sincronización entre dispositivos de las sesiones de chat. También incluye herramientas para el filtrado de contenido y la personalización de prompts por sesión para garantizar una salida consistente y segura.
El software está diseñado para su despliegue en contenedores, asegurando una ejecución consistente en diversos entornos de infraestructura. Puede instalarse como una aplicación web progresiva (PWA) para proporcionar una experiencia nativa tanto en dispositivos móviles como de escritorio.
Features
- AI Model Interfaces - Provides a web-based chat interface for interacting with large language models.
- Self-Hosted AI Infrastructure - Provides a self-hosted, containerized chat platform for managing AI interactions and conversation history on private infrastructure.
- LLM Chat Interfaces - Offers a browser-based interface for real-time streaming communication with multiple large language models.
- Chat History Synchronization - Persists and synchronizes chat session history across multiple devices to ensure continuity of conversations.
- Agent Session Persistence - Persists conversation history and user context in a database to enable cross-device session resumption.
- Container Deployment - Supports containerized deployment for consistent execution across diverse infrastructure environments.
- API Key Management - Manages multiple API keys with randomized selection to ensure continuous access to backend model providers.
- Password Access Restrictions - Protects the service by requiring password-based authentication to prevent unauthorized access to the backend API.
- Enterprise AI Security - Secures access to AI models through enterprise identity provider integration and usage quota enforcement.
- Conversational Session Managers - Organizes conversation history with tools for managing, exporting, and editing chat sessions.
- Server-Side Resource Proxies - Routes requests through a backend server to mask API keys and bypass cross-origin restrictions.
- Persona and Behavioral Instructions - Allows assigning unique system instructions to individual chat sessions to tailor AI behavior.
- Context Window Management - Allows users to adjust conversation history limits to balance model performance and memory usage.
- OpenAI API Clients - Acts as a web-based client for OpenAI-compatible APIs, supporting prompt templates and API key management.
- Prompt Templates - Enables the import and organization of custom prompt libraries for consistent AI interactions.
- Real-Time Web Search Integrations - Integrates real-time web search capabilities to ground AI responses with current information.
- Reasoning Depth Configurations - Provides configuration toggles to adjust the reasoning depth and complexity of compatible language models.
- Proxy Routing - Routes API requests through custom network proxies to bypass regional restrictions and manage secure connectivity.
- Containerized Deployments - Packages the application into portable container images for consistent deployment across diverse infrastructure environments.
- Request Rate Limiting - Monitors and restricts incoming request frequency based on IP addresses to prevent service abuse.
- Self-Hosted AI Clients - Provides a centralized dashboard for managing AI model interactions and user authentication in a self-hosted environment.
- Client Side Rendering - Uses a reactive framework to dynamically render rich content like code blocks and mathematical formulas.
- AI Content Filters - Filters user inputs and model responses using custom sensitive word lists to maintain safety and compliance.
- Enterprise Identity Providers - Integrates with enterprise identity providers for centralized authentication and access control.
- Proxy-based Access Controls - Validates user identities by inspecting verified headers passed from an external reverse proxy.
- User Account Management - Provides comprehensive user account management including registration, login, and administrative access controls.
- AI Content Renderers - Renders streamed AI-generated content including markdown, code blocks, and mathematical notation directly in the chat interface.
- Response Streaming - Handles real-time streaming of model responses to display text incrementally as it is generated.
