# LanceDB MCP for AI Agents AI Agent Connect > LanceDB MCP connects your LanceDB Cloud account to your AI agent. It enables similarity searches, table creation, and real-time data ingestion for multi-modal embeddings. It turns your database management into a natural conversation, allowing you to handle vector orchestration and RAG infrastructure without writing manual scripts. ## Overview - **Category:** loved-by-devs - **Price:** Free - **Endpoint:** https://edge.vinkius.com/vk_preview_85lSO635ajE8lXHCZKWQ7UC6VHfybBdy20nU97wm/ai-agent-connect - **Tags:** vector-search, embeddings, rag, serverless-db, multi-modal, data-indexing ## Description This Connector connects your LanceDB Cloud account to your AI agent so you can manage your cloud-hosted vector storage and RAG infrastructure through natural conversation. Instead of jumping between your database console and your code editor, you can handle the heavy lifting of vector orchestration without manually writing Python scripts every time you need to check a schema or run a search. You can manage RAG systems and complex multi-modal embeddings by asking your agent to provision new tables with specific Apache Arrow schemas or pull the latest metadata on your tensor dimensions. It handles the messy parts of database maintenance, like cleaning up old tables or verifying storage configurations. You can find this Connector in the Vinkius catalog to get your vector storage connected to your preferred client in minutes. It turns your agent into a data engineer that understands your vector topologies and can execute high-speed KNN lookups on demand. You no longer have to worry about the technical overhead of indexing or schema alignment. Your agent takes on the responsibility of maintaining the integrity of your vector data while you focus on building the actual application logic. ## Tools ### list_tables See every vectorized table currently living in your LanceDB instance. This helps you keep track of your active data collections. ### get_table Pull the exact schema and metadata for a specific table to check tensor dimensions. It ensures your agent knows the data structure. ### vector_search Execute a high-speed KNN similarity search to find semantically related rows. This is the primary way to perform RAG lookups. ### insert_rows Add new structured data and vectors while the system updates the ANN index automatically. It keeps your search results current. ### create_table Provision a new vector table with a custom Apache Arrow schema for your specific workload. This ensures strict data integrity. ### delete_table Permanently remove a vector table to keep your storage environment clean. Use this to vaporize old test data. ## Prompt Examples **Prompt:** ``` List all the tables in my LanceDB instance. ``` **Response:** ``` I've retrieved your vector tables. Active tables include 'customer_docs', 'product_embeddings', 'support_kb', and 'user_logs'. Which one would you like to inspect the schema for? ``` **Prompt:** ``` Perform a vector search in 'product_embeddings' for this vector: [0.1, 0.2, ...] ``` **Response:** ``` Executing similarity search... I've found the 5 most relevant products. Highlights include 'Cloud-Native Sneakers' (Distance: 0.12), 'Edge Performance Runner' (Distance: 0.15), and 'Vector Pro Trainer' (Distance: 0.18). All results are now available in your workspace. ``` **Prompt:** ``` Show me the schema for the 'support_kb' table ``` **Response:** ``` Schema for 'support_kb': The table has 4 columns: 'id' (String), 'text' (String), 'vector' (Float32, Dimensions: 1536), and 'metadata' (JSON). It is currently using an IVF-PQ index for optimized ANN lookups. ``` ## Capabilities ### Run KNN similarity searches Execute high-speed lookups to find semantically related rows in your vector tables. ### Create new vector tables Provision new tables with specific Apache Arrow schemas for your AI workloads. ### Insert row payloads Add new data and vectors while the system updates the ANN index in real time. ### Fetch table schemas Pull exact metadata and tensor dimensions for any specific table. ### List all vectorized tables See a full list of all active tables residing in your LanceDB instance. ### Delete vector tables Permanently remove entire tables to keep your storage environment clean. ## Use Cases ### Verifying document retrieval A developer wants to see if a new document chunk was indexed correctly. They ask the agent to list all tables and then check the schema of the specific collection to verify the tensor dimensions. ### Provisioning new RAG pipelines A data engineer needs to set up a new RAG pipeline for a client. They tell the agent to provision a new table with a specific float32 dimension for their embeddings to ensure it meets the project requirements. ### Multi-modal image search A researcher wants to find similar images in a multi-modal set. They provide a vector and ask the agent to perform a KNN search on the product table to find the most relevant visual matches. ### Auditing storage costs An architect needs to audit their cloud storage usage and limits. They ask the agent to find all active tables and verify the storage configurations across their various database instances. ## Benefits - Stop writing manual Python scripts for similarity searches by using vector_search to get results instantly. This lets you iterate on your retrieval logic in seconds rather than minutes. - Ensure data consistency with create_table, which lets you define strict Apache Arrow schemas for your AI workloads. This prevents schema mismatches that often break production pipelines. - Keep your database organized by using delete_table to purge old data and maintain a clean storage environment. It helps you manage costs and keep your active data sets focused. - Get immediate feedback on your data structure by using get_table to check dimensions and topologies. You can verify your embeddings are correct before you deploy them to your users. - Simplify multi-modal data ingestion with insert_rows, which updates your ANN index in real time. This ensures your agent always has access to the most recent information without manual refreshes. - Monitor your entire infrastructure at a glance by using list_tables to see all active vectorized tables. It gives you a clear view of your storage footprint across different projects. ## How It Works The bottom line is you get a direct conversational interface for your cloud-hosted vector database. 1. Connect your LanceDB Cloud credentials to the Connector. 2. Specify your API URL, API Key, and Database Name. 3. Ask your agent to perform searches or manage tables. ## Frequently Asked Questions **What does the LanceDB MCP do for my RAG system?** It gives your AI agent the ability to interact directly with your vector storage. You can ask it to perform similarity searches, check table schemas, or manage your data without writing any extra code. **Can I use LanceDB MCP to search through my embeddings?** Yes, you can perform high-speed KNN similarity searches. Just provide the vector to your agent, and it will find the most relevant rows in your LanceDB instance. **How does the LanceDB MCP handle new data?** When you insert new rows, the Connector ensures the underlying ANN index is updated in real time. This means your agent always sees the most current information. **Can I create new tables using the LanceDB MCP?** Absolutely. You can ask your agent to provision new vector tables with specific Apache Arrow schemas to ensure your data remains consistent and organized. **How do I manage my LanceDB Cloud storage with this Connector?** You can manage your storage by listing all active tables, verifying their configurations, and deleting old tables to keep your environment clean and optimized. **Does the LanceDB MCP support multi-modal data?** Yes, it is designed to handle multi-modal embeddings. You can manage the different topologies and schemas required for complex AI workloads through natural conversation. **Can I perform a semantic similarity search using my agent?** Yes. Use the `vector_search` tool by providing the target Table name and a JSON array of floating-point numbers representing your query embedding. Your agent will return the k-nearest rows from LanceDB based on semantic similarity. **How do I create a new table with a specific Apache Arrow schema?** The `create_table` tool allows your agent to initialize a new columnar vector table. You just need to provide the desired Table name and a valid Apache Arrow schema mapping in JSON format defining dimensions and scalar fields. **Can my agent insert new embeddings directly into a LanceDB table?** Absolutely. Use the `insert_rows` tool to persist new data rows containing native embeddings and arbitrary JSON metadata. Your agent will handle the payload delivery, and LanceDB will automatically update its ANN index.