110 providers. 19 categories.
Every provider registers itself in init(). Import for
side-effect, call <pkg>.New — middleware, hooks,
and metrics attach automatically. Search by name or filter by
category to find the one you need.
Package anthropic provides the Anthropic (Claude) LLM provider for the Beluga AI framework.
Package azure provides the Azure OpenAI LLM provider for the Beluga AI framework.
Package bedrock provides the AWS Bedrock LLM provider for the Beluga AI framework.
Package bifrost provides a ChatModel backed by a Bifrost gateway for the Beluga AI framework.
Package cerebras provides the Cerebras LLM provider for the Beluga AI framework.
Package cohere provides the Cohere LLM provider for the Beluga AI framework.
Package deepseek provides the DeepSeek LLM provider for the Beluga AI framework.
Package fireworks provides the Fireworks AI LLM provider for the Beluga AI framework.
Package google provides the Google Gemini LLM provider for the Beluga AI framework.
Package groq provides the Groq LLM provider for the Beluga AI framework.
Package huggingface provides the HuggingFace Inference API LLM provider for the Beluga AI framework.
Package litellm provides a ChatModel backed by a LiteLLM gateway for the Beluga AI framework.
Package llama provides a Meta Llama model provider for the Beluga AI framework.
Package mistral provides the Mistral AI LLM provider for the Beluga AI framework.
Package ollama provides the Ollama LLM provider for the Beluga AI framework.
Package openai provides the OpenAI LLM provider for the Beluga AI framework.
Package openrouter provides the OpenRouter LLM provider for the Beluga AI framework.
Package perplexity provides the Perplexity LLM provider for the Beluga AI framework.
Package qwen provides the Alibaba Qwen LLM provider for the Beluga AI framework.
Package sambanova provides the SambaNova LLM provider for the Beluga AI framework.
Package together provides the Together AI LLM provider for the Beluga AI framework.
Package xai provides the xAI Grok LLM provider for the Beluga AI framework.
Package chroma provides a VectorStore backed by ChromaDB.
Package elasticsearch provides a VectorStore backed by Elasticsearch's kNN search.
Package inmemory provides an in-memory VectorStore for testing and small-scale use.
Package milvus provides a VectorStore backed by the Milvus vector database.
Package mongodb provides a VectorStore backed by MongoDB Atlas Vector Search.
Package pgvector provides a VectorStore backed by PostgreSQL with the pgvector extension.
Package pinecone provides a VectorStore backed by the Pinecone vector database.
Package qdrant provides a VectorStore backed by the Qdrant vector database.
Package redis provides a VectorStore backed by Redis with the RediSearch module.
Beluga AI provider for sqlitevec.
Package turbopuffer provides a VectorStore backed by the Turbopuffer serverless vector database.
Package vespa provides a VectorStore backed by the Vespa search engine.
Package weaviate provides a VectorStore backed by the Weaviate vector database.
Package cohere provides a Cohere embeddings provider for the Beluga AI framework.
Package google provides a Google AI embeddings provider for the Beluga AI framework.
Package inmemory provides a deterministic hash-based Embedder for testing.
Package jina provides a Jina AI embeddings provider for the Beluga AI framework.
Package mistral provides an Embedder backed by the Mistral AI embeddings API.
Package ollama provides an Ollama embeddings provider for the Beluga AI framework.
Package openai provides an OpenAI embeddings provider for the Beluga AI framework.
Package sentencetransformers provides an Embedder backed by the HuggingFace Inference API for Sentence Transformers models.
Package voyage provides a Voyage AI embeddings provider for the Beluga AI framework.
Package cloudstorage provides a DocumentLoader that loads files from cloud storage services (S3, GCS, Azure Blob).
Package confluence provides a DocumentLoader that loads pages from Atlassian Confluence via its REST API.
Package docling provides a DocumentLoader that uses the IBM Docling API to convert documents (PDFs, DOCX, images, etc.) into structured content.
Package firecrawl provides a DocumentLoader that uses Firecrawl to crawl websites and extract their content as markdown.
Package gdrive provides a DocumentLoader that loads files from Google Drive via the Google Drive REST API.
Package github provides a DocumentLoader that loads files from GitHub repositories via the GitHub API.
Package notion provides a DocumentLoader that loads pages from Notion via its API.
Package unstructured provides a DocumentLoader that uses the Unstructured.io API to extract structured content from files (PDFs, DOCX, images, etc.).
Package assemblyai provides the AssemblyAI STT provider for the Beluga AI voice pipeline.
Package deepgram provides the Deepgram STT provider for the Beluga AI voice pipeline.
Package elevenlabs provides the ElevenLabs Scribe STT provider for the Beluga AI voice pipeline.
Package gladia provides the Gladia STT provider for the Beluga AI voice pipeline.
Package groq provides the Groq STT provider for the Beluga AI voice pipeline.
Package whisper provides the OpenAI Whisper STT provider for the Beluga AI voice pipeline.
Package cartesia provides the Cartesia TTS provider for the Beluga AI voice pipeline.
Package elevenlabs provides the ElevenLabs TTS provider for the Beluga AI voice pipeline.
Package fish provides the Fish Audio TTS provider for the Beluga AI voice pipeline.
Package groq provides the Groq TTS provider for the Beluga AI voice pipeline.
Package lmnt provides the LMNT TTS provider for the Beluga AI voice pipeline.
Package playht provides the PlayHT TTS provider for the Beluga AI voice pipeline.
Package smallest provides the Smallest.ai TTS provider for the Beluga AI voice pipeline.
Package gemini provides the Gemini Live S2S provider for the Beluga AI voice pipeline.
Package nova provides the Amazon Nova S2S provider for the Beluga AI voice pipeline.
Package openai provides the OpenAI Realtime S2S provider for the Beluga AI voice pipeline.
Package daily provides the Daily.co transport provider for the Beluga AI voice pipeline.
Package livekit provides the LiveKit transport provider for the Beluga AI voice pipeline.
Package pipecat provides the Pipecat transport provider for the Beluga AI voice pipeline.
Beluga AI provider for silero.
Package webrtc provides a pure Go WebRTC-style VAD (Voice Activity Detection) provider for the Beluga AI voice pipeline.
Package azuresafety provides an Azure Content Safety guard implementation for the Beluga AI safety pipeline.
Package guardrailsai provides a Guardrails AI guard implementation for the Beluga AI safety pipeline.
Package lakera provides a Lakera Guard API guard implementation for the Beluga AI safety pipeline.
Package llmguard provides an LLM Guard API guard implementation for the Beluga AI safety pipeline.
Package nemo provides an NVIDIA NeMo Guardrails guard implementation for the Beluga AI safety pipeline.
Package dapr provides a Dapr state store-backed [workflow.WorkflowStore] implementation for the Beluga AI workflow engine.
Package inmemory provides an in-memory [workflow.WorkflowStore] for development and testing.
Package inngest provides an Inngest-backed [workflow.WorkflowStore] implementation for the Beluga AI workflow engine.
Package kafka provides a Kafka-backed [workflow.WorkflowStore] implementation for the Beluga AI workflow engine.
Package nats provides a NATS JetStream KV-backed [workflow.WorkflowStore] implementation for durable workflow state persistence.
Package temporal provides a Temporal-backed [workflow.DurableExecutor] and [workflow.WorkflowStore] for the Beluga workflow engine.
Package braintrust provides a Braintrust evaluation metric for the Beluga AI eval framework.
Package deepeval provides a DeepEval evaluation metric for the Beluga AI eval framework.
Package ragas provides RAGAS (Retrieval Augmented Generation Assessment) evaluation metrics for the Beluga AI eval framework.
Package inmemory provides an in-memory implementation of the [state.Store] interface.
Package inmemory provides an in-memory LRU cache implementation for the Beluga AI framework.
Package file provides a filesystem-based PromptManager that loads prompt templates from a directory of JSON files.
Package composio provides a Composio MCP integration for the Beluga AI protocol layer.
Package chi provides a Chi-based ServerAdapter for the Beluga AI server package.
Package connect provides a Connect-Go based ServerAdapter for the Beluga AI server package.
Package echo provides an Echo-based ServerAdapter for the Beluga AI server package.
Package fiber provides a Fiber v3-based ServerAdapter for the Beluga AI server package.
Package gin provides a Gin-based ServerAdapter for the Beluga AI server package.
Package grpc provides a gRPC-based ServerAdapter for the Beluga AI server package.
Package huma provides a Huma-based ServerAdapter for the Beluga AI server package.
Package dragonfly provides a DragonflyDB-backed implementation of [memory.MessageStore].
Package inmemory provides in-memory implementations of the memory store interfaces.
Package langfuse provides a Langfuse trace exporter for the Beluga AI observability system.
Package langsmith provides a LangSmith trace exporter for the Beluga AI observability system.
Package memgraph provides a Memgraph-backed [memory.GraphStore] implementation for the Beluga AI memory system.
Package mongodb provides a MongoDB-backed implementation of [memory.MessageStore].
Package neo4j provides a Neo4j-backed [memory.GraphStore] implementation for the Beluga AI memory system.
Package opik provides an Opik trace exporter for the Beluga AI observability system.
Package phoenix provides an Arize Phoenix trace exporter for the Beluga AI observability system.
Package postgres provides a PostgreSQL-backed implementation of [memory.MessageStore].
Package redis provides a Redis-backed implementation of [memory.MessageStore].
Package sqlite provides a SQLite-backed implementation of [memory.MessageStore].
No providers match. Try a broader search or clear the category filter.
Adding your own?
The four-ring extensibility pattern applies identically to every provider category. Interface, registry, hooks, middleware — once you know the shape, every package is learnable in minutes.
Read the provider template →