by Voyage AI
Voyage AI provides embedding models and rerankers through API endpoints. They are used for semantic search and retrieval-augmented generation.
APIUsage-based
Step 1: Sign in to the Voyage dashboard and create a secret key in the API keys section.
Step 2: Set the key as the VOYAGE_API_KEY environment variable.
Step 3: Install the client with pip install -U voyageai.
Step 4: Create a voyageai.Client and call embed on your documents with input_type set to document.
Step 5: Embed the user query with input_type set to query and compare it to the document vectors.
Step 6: Optionally rerank the top results with a reranker.
Embedding models and rerankers, available through API endpoints.
Create an API key in the dashboard and set VOYAGE_API_KEY, or pass api_key to the client.
No. The API works over HTTP from any language. A Python package is provided for convenience.
The quickstart batches in groups of 128 and notes per-request token and context limits, which you can check with the tokenizer and count_tokens.
FirecrawlFirecrawl is a web data API that searches, scrapes and interacts with web pages and returns clean, LLM-ready data. It is open source and works with AI agents through MCP and a CLI.
ExaExa is a search API for AI agents that returns web data, page contents and structured research results. It offers search, contents, agent and monitor endpoints.