About this project

Wynth API is a downstream fork of New API designed as a next-generation LLM gateway and AI asset management system. It provides a centralized interface for managing multiple AI model providers, offering organization-level authentication, usage analytics, and cost accounting. Key capabilities include: - Account Pooling: Multi-platform account pooling for OpenAI, Anthropic, Gemini, Vertex AI, xAI, and grok.com, featuring failure-grading cooldowns, load-aware selection, and per-user concurrency controls. - Model Support: Integration with OpenAI-compatible models, Claude (Messages format), Google Gemini, Midjourney-Proxy, Suno-API, and Rerank models (Cohere, Jina). - Intelligent Routing: Supports weighted random channel selection, automatic retries on failure, and user-level rate limiting. - Format Conversion: Translates between OpenAI compatible, Claude Messages, and Google Gemini formats. - Management Tools: A modern UI with a data dashboard, permission management (token grouping, model restrictions), and support for various authentication methods including OIDC, Discord, Telegram, and LinuxDO. - Deployment: Supports Docker and Docker Compose, with database options for SQLite, MySQL, and PostgreSQL.