canonical: https://jentic.com/apis/meilisearch.com/meilisearch-v1-0 # Meilisearch v1.0 Meilisearch is a self-hosted search engine you run on your own instance. This API creates and configures indexes, adds and updates documents, and runs fast full-text search with typo tolerance, filtering, faceting, and custom ranking rules. It exposes fine-grained index settings such as searchable and displayed attributes, synonyms, stop-words, and sortable attributes, manages API keys, and tracks the asynchronous tasks that document and settings changes create. ## For AI agents Create and configure search indexes, add and update documents, and run fast full-text search with typo tolerance, filtering, and faceting on a self-hosted Meilisearch instance. Manage index settings, API keys, and asynchronous tasks. Returns ranked search hits and task status. ## Scope Provides full-text search, indexing, and index configuration on a self-hosted instance; does not host a language-model or chat endpoint or manage hosting and billing through this surface. Use for search and indexing. ## Capabilities - Create, list, update, and delete search indexes - Add, replace, update, and delete documents in an index - Run full-text search with typo tolerance, filtering, and faceting - Configure index settings such as searchable attributes, synonyms, and ranking rules - Create, list, update, and delete API keys with scoped permissions - Track the asynchronous tasks that document and settings changes create ## Use cases ### Search backend for an AI agent Give an AI agent a search layer it can populate and query on a Meilisearch instance you run yourself. Through Jentic the agent discovers the index, document, and search operations by intent and calls them with a validated input schema, so an assistant can index a corpus and then run full-text search to fetch the most relevant records for a question, with typo tolerance handling imperfect queries. Example prompt: Create an index, add a batch of help articles as documents, then run a full-text search for a user's question and return the top three hits ### Site and app search Power an instant search experience for a website or application, returning ranked hits with typo tolerance, filtering, and faceting. You configure searchable attributes, synonyms, and ranking rules so results match how your users actually query, and the search endpoint returns results fast enough for as-you-type interfaces. Example prompt: Search the products index for 'wireles headphnes' and return the corrected, ranked results using typo tolerance ### Index configuration and content sync Keep an index aligned with a source of truth by adding, updating, and deleting documents, and tune relevance through index settings such as searchable attributes, synonyms, stop-words, and ranking rules. Because changes are asynchronous, you can track each task to confirm when an update has been applied. Example prompt: Update the synonyms setting on the products index and then check the resulting task until it reports success ### Scoped access with API keys Create and manage API keys with scoped permissions so different clients get only the access they need, such as a search-only key for a public front end and an admin key for indexing. Listing and updating keys lets you rotate and audit access to your Meilisearch instance. Example prompt: Create a search-only API key restricted to the products index for use in a public web front end ## Key endpoints | Method | Path | Description | | --- | --- | --- | | POST | `/indexes` | Create an index | | GET | `/indexes` | List indexes | | POST | `/indexes/books/documents` | Add or replace documents in an index (books is the spec's example index) | | POST | `/indexes/books/search` | Search documents in an index (books is the spec's example index) | | GET | `/tasks` | List asynchronous tasks | | GET | `/keys` | List API keys | ## Key resources - **Indexes** — Create, list, update, and delete search indexes - **Documents** — Add, replace, update, fetch, and delete documents in an index - **Search** — Full-text search with typo tolerance, filtering, and faceting - **Settings** — Configure searchable attributes, synonyms, stop-words, ranking rules, and more - **Keys** — Create and manage API keys with scoped permissions - **Tasks** — Track the asynchronous tasks created by index and document changes ## Why Jentic - **Setup:** Wiring Meilisearch by hand means pointing at your own instance, sending the API key as a bearer token on every request, and tracking the asynchronous task each change returns. Through Jentic you install once, import Meilisearch from the API Directory, store the key once, and your agent calls the index, document, and search operations directly. - **Permission scoping:** You choose the operations your agent may call, so you can allow searching and reading documents while withholding index and document deletion or key management. That way a search assistant can query your instance without the ability to destroy indexes or mint new keys. - **Credential handling:** Your Meilisearch API key is stored once, encrypted, by your own Jentic One instance and injected at execution time. It never enters the agent's prompt, logs, or context, and because the instance is self-hosted your search data stays on infrastructure you control. - **Discovery method:** Agents search Jentic by intent such as 'run a full-text search' or 'add documents to an index', and Jentic returns the matching Meilisearch operation with its input schema so the agent calls the right endpoint without reading the reference docs. ## Related APIs - **Algolia Search API** — Hosted keyword search tuned for instant, typo-tolerant results. - **Marqo Cloud API** — Vector search that ranks by semantic similarity with built-in embeddings. - **Pinecone** — Managed vector database where you supply your own embeddings. ## FAQ ### What authentication does the Meilisearch API use? Meilisearch authenticates with an API key sent as a bearer token in the Authorization header. You manage keys through the keys endpoints, creating scoped keys for search-only or admin access. Through Jentic the key is stored once and injected at call time, so your agent never places it in a header itself. ### Is Meilisearch self-hosted, and what does that mean for my agent? Yes. Meilisearch runs on your own instance rather than a shared cloud endpoint, so your documents and search traffic stay on infrastructure you control. Through Jentic your agent points at your instance and calls the index, document, and search operations without exposing the instance or its key in the prompt. ### Is there a Meilisearch MCP server? You do not need an MCP server to give your agent Meilisearch. Jentic connects it directly from the API Directory: import it, store your API key once, and your agent calls the index, document, and search operations on demand. Nothing extra runs, and no server tool definitions sit in the agent's context between calls. ### Can I limit what my agent is allowed to do with Meilisearch? Yes. Write a rule that allows only the operations you need, such as running searches while leaving out index and document deletion, so the agent can query your data but cannot destroy it. Every call it makes is logged, so you can audit exactly which operations ran. ### What are the rate limits for Meilisearch? Because Meilisearch is self-hosted, there is no vendor-imposed rate limit. Throughput is bounded only by the resources of the instance you run it on, so you scale it to your own workload rather than a plan tier. ### How do I run a search with Meilisearch through Jentic? Search Jentic for 'run a full-text search' to find the search operation, which returns ranked hits from an index with typo tolerance, and pair it with the add-documents operation to populate the index first. Jentic returns each operation with its input schema so your agent calls it correctly. To run it on your own infrastructure, install Jentic One from its GitHub repo.