Vibe Coding Discoveresperanto
View on GitHubA unified interface for various AI model providers
A lightweight Python library with a consistent API for LLMs, embeddings, rerankers, transcription, and speech generation across 15+ hosted and local providers. Includes async and streaming support, structured outputs, and optional LangChain integration.
Use Cases
Build Python applications that switch between LLM providers through one APIGenerate text with synchronous, asynchronous, or streaming callsCreate embeddings for semantic search and retrievalRerank search resultsTranscribe audio across supported providersGenerate speech from textRun local models through Ollama, vLLM, or TransformersIntegrate supported models with LangChain
Built With
- Language
- Python
- Frameworks
- LangChain · Hugging Face Transformers · Sentence Transformers · Ollama · vLLM
Tags
LLM · multi-provider · unified API · embeddings · reranking · speech-to-text · text-to-speech · streaming · async · structured output · model discovery · local inference