Weaviate MCP Server for Cloudflare Workers
Test Weaviate tools directly in your browser. Select a tool and provide inputs to see the results.
Create a free account to test Weaviate tools interactively. No credit card required.
weaviate_create_backupCreate a backup of Weaviate data. Backups can be stored in different backends (filesystem, s3, gcs, azure). Args: - backend: Backup backend (filesystem, s3, gcs, azure) - backupId: Unique identifier for this backup - include: Array of class names to include (optional, default: all) - exclude: Array of class names to exclude (optional) Returns the backup status.
weaviate_get_backup_statusGet the status of a backup operation. Args: - backend: Backup backend (filesystem, s3, gcs, azure) - backupId: Backup identifier Returns the current backup status.
weaviate_restore_backupRestore data from a backup. Args: - backend: Backup backend (filesystem, s3, gcs, azure) - backupId: Backup identifier to restore from - include: Array of class names to include in restore (optional, default: all) - exclude: Array of class names to exclude from restore (optional) Returns the restore operation status.
weaviate_get_restore_statusGet the status of a restore operation. Args: - backend: Backup backend (filesystem, s3, gcs, azure) - backupId: Backup identifier Returns the current restore status.
weaviate_cancel_backupCancel an in-progress backup or restore operation. Args: - backend: Backup backend (filesystem, s3, gcs, azure) - backupId: Backup identifier to cancel Returns confirmation of cancellation.
weaviate_deactivate_tenantsDeactivate (cool down) tenants to reduce resource usage. Deactivated tenants are still stored but not loaded into memory. This is a convenience method that sets tenant status to INACTIVE. Args: - className: Name of the multi-tenant class - tenantNames: Array of tenant names to deactivate Returns confirmation of deactivation.
weaviate_batch_createCreate multiple objects in a single batch request. This is the most efficient way to import large amounts of data. Args: - objects: Array of objects to create, each with: - class: Class name - properties: Object properties - id: Optional UUID - vector: Optional custom vector - tenant: Optional tenant name Returns the batch operation results.
weaviate_batch_deleteDelete multiple objects matching a filter in a single batch request. Args: - className: Name of the class - where: Where filter to match objects for deletion - dryRun: If true, only report how many would be deleted without actually deleting - output: Output verbosity ('minimal' or 'verbose') Returns the deletion results.
weaviate_batch_add_referencesAdd multiple cross-references in a single batch request. Args: - references: Array of references to add, each with: - from: Source beacon (weaviate://localhost/ClassName/uuid/propertyName) - to: Target beacon (weaviate://localhost/ClassName/uuid) Returns the batch operation results.
weaviate_create_classificationStart a classification task to automatically classify objects. Classification types: - knn: k-nearest neighbors based on vector similarity - text2vec-contextionary-contextual: Contextual classification using contextionary - zeroshot: Zero-shot classification Args: - type: Classification type (knn, text2vec-contextionary-contextual, zeroshot) - className: Class to classify objects in - classifyProperties: Properties to classify (reference properties) - basedOnProperties: Properties to base classification on - k: Number of neighbors for knn (default: 3) - sourceWhere: Filter for source objects to classify - targetWhere: Filter for target objects to use as references - trainingSetWhere: Filter for training set (knn) Returns the classification task status.
weaviate_get_classificationGet the status of a classification task. Args: - id: Classification task ID Returns the classification status and results.
weaviate_get_metaGet Weaviate server metadata including version and modules. Returns server information including hostname, version, and configured modules.
weaviate_is_liveCheck if Weaviate is alive (liveness probe). This is a simple health check that returns ok if the server is running.
weaviate_is_readyCheck if Weaviate is ready to accept requests (readiness probe). This checks if the server is fully initialized and ready for queries.
weaviate_get_nodesGet information about all nodes in the Weaviate cluster. Returns status, version, and statistics for each node. Args: - output: Output verbosity (minimal or verbose) Returns list of nodes with their status.
weaviate_get_cluster_statisticsGet cluster-wide statistics. Returns aggregate statistics across all nodes in the cluster.
weaviate_list_objectsList objects from a Weaviate class with pagination. Args: - className: Name of the class - limit: Number of objects to return (1-100, default: 20) - offset: Offset for pagination - include: Additional fields to include (vector, classification, featureProjection) - tenant: Tenant name for multi-tenant collections Returns a paginated list of objects.
weaviate_get_objectGet a single object by ID. Args: - className: Name of the class - id: Object ID (UUID) - include: Additional fields to include (vector, classification, featureProjection) - tenant: Tenant name for multi-tenant collections Returns the object with all its properties.
weaviate_create_objectCreate a new object in a Weaviate class. Args: - className: Name of the class - properties: Object properties as key-value pairs - id: Optional UUID for the object (auto-generated if not provided) - vector: Optional custom vector (required if class has no vectorizer) - tenant: Tenant name for multi-tenant collections Returns the created object.
weaviate_update_objectUpdate an existing object (replaces all properties). Args: - className: Name of the class - id: Object ID (UUID) - properties: New object properties (replaces all existing) - vector: Optional new vector - tenant: Tenant name for multi-tenant collections Returns the updated object.
weaviate_patch_objectPartially update an object (only provided properties are updated). Args: - className: Name of the class - id: Object ID (UUID) - properties: Properties to update (existing properties not included are preserved) - tenant: Tenant name for multi-tenant collections Returns confirmation of update.
weaviate_delete_objectDelete an object by ID. Args: - className: Name of the class - id: Object ID (UUID) to delete - tenant: Tenant name for multi-tenant collections Returns confirmation of deletion.
weaviate_object_existsCheck if an object exists by ID (HEAD request). Args: - className: Name of the class - id: Object ID (UUID) to check - tenant: Tenant name for multi-tenant collections Returns whether the object exists.
weaviate_add_referenceAdd a cross-reference from one object to another. Cross-references link objects together. The source object must have a reference property defined in its class schema. Args: - className: Source class name - id: Source object ID - propertyName: Name of the reference property - targetClass: Target class name - targetId: Target object ID - tenant: Tenant name for multi-tenant collections Returns confirmation of reference addition.
weaviate_update_referencesReplace all references for a property with a new set. This replaces ALL existing references on the property with the provided list. Args: - className: Source class name - id: Source object ID - propertyName: Name of the reference property - targets: Array of target references with class and id - tenant: Tenant name for multi-tenant collections Returns confirmation of reference update.
weaviate_delete_referenceDelete a specific cross-reference from an object. Args: - className: Source class name - id: Source object ID - propertyName: Name of the reference property - targetClass: Target class name - targetId: Target object ID to remove - tenant: Tenant name for multi-tenant collections Returns confirmation of reference deletion.
weaviate_get_schemaGet the full Weaviate schema with all collections/classes. Returns the complete schema including all classes, their properties, vectorizers, and configurations.
weaviate_get_classGet a specific class/collection from the schema. Args: - className: The name of the class to retrieve Returns the class configuration including properties, vectorizer, and module config.
weaviate_create_classCreate a new class/collection in Weaviate. Args: - className: Name of the class (must start with uppercase) - description: Description of the class - vectorizer: Vectorizer module (text2vec-openai, text2vec-cohere, text2vec-huggingface, none) - properties: Array of property definitions - moduleConfig: Module-specific configuration - vectorIndexType: Type of vector index (hnsw, flat) - multiTenancyEnabled: Enable multi-tenancy for this class Returns the created class configuration.
weaviate_update_classUpdate an existing class/collection configuration. Note: Not all properties can be updated after creation. Typically you can update: - description - invertedIndexConfig - replicationConfig - vectorIndexConfig (some settings) Args: - className: Name of the class to update - description: New description - invertedIndexConfig: Inverted index configuration - replicationConfig: Replication configuration Returns the updated class configuration.
weaviate_delete_classDelete a class/collection and all its data. WARNING: This permanently deletes the class and all objects in it. Args: - className: Name of the class to delete Returns confirmation of deletion.
weaviate_add_propertyAdd a new property to an existing class. Args: - className: Name of the class - name: Property name - dataType: Array of data types (e.g., ["text"], ["int"], ["boolean"]) - description: Property description - indexFilterable: Enable filtering on this property - indexSearchable: Enable full-text search - tokenization: Tokenization method (word, lowercase, whitespace, field) Returns confirmation of property addition.
weaviate_graphqlExecute a raw GraphQL query against Weaviate. Use this for complex queries not covered by other search tools. Args: - query: The GraphQL query string Returns the GraphQL response.
weaviate_near_vectorSearch for objects similar to a given vector. Args: - className: Name of the class to search - vector: The query vector (array of numbers) - limit: Maximum number of results (default: 10) - certainty: Minimum certainty threshold (0-1) - distance: Maximum distance threshold - fields: Fields to return (default: _additional { id distance certainty }) - where: Optional filter - tenant: Tenant name for multi-tenant collections Returns matching objects sorted by similarity.
weaviate_near_textSearch for objects semantically similar to text concepts. Requires a text vectorizer module (text2vec-openai, text2vec-cohere, etc.) Args: - className: Name of the class to search - concepts: Array of text concepts to search for - limit: Maximum number of results (default: 10) - certainty: Minimum certainty threshold (0-1) - distance: Maximum distance threshold - moveTo: Move results towards these concepts (optional) - moveAwayFrom: Move results away from these concepts (optional) - fields: Fields to return - where: Optional filter - tenant: Tenant name for multi-tenant collections Returns matching objects sorted by semantic similarity.
weaviate_near_objectSearch for objects similar to an existing object by ID. Args: - className: Name of the class to search - id: Object ID to find similar objects to - limit: Maximum number of results (default: 10) - certainty: Minimum certainty threshold (0-1) - distance: Maximum distance threshold - fields: Fields to return - where: Optional filter - tenant: Tenant name for multi-tenant collections Returns matching objects sorted by similarity.
weaviate_hybridPerform hybrid search combining vector search with keyword search (BM25). Args: - className: Name of the class to search - query: The search query text - alpha: Balance between vector (1) and keyword (0) search (default: 0.5) - vector: Optional custom vector (if not using text vectorizer) - properties: Properties to search in for BM25 - fusionType: How to fuse results (rankedFusion or relativeScoreFusion) - limit: Maximum number of results (default: 10) - fields: Fields to return - where: Optional filter - tenant: Tenant name for multi-tenant collections Returns matching objects combining both search methods.
weaviate_bm25Perform BM25 keyword search (no vector search). Args: - className: Name of the class to search - query: The search query text - properties: Properties to search in - limit: Maximum number of results (default: 10) - fields: Fields to return - where: Optional filter - tenant: Tenant name for multi-tenant collections Returns matching objects based on keyword relevance.
weaviate_get_tenantsList all tenants for a multi-tenant class. Args: - className: Name of the multi-tenant class Returns list of tenants with their activity status.
weaviate_create_tenantsCreate one or more tenants for a multi-tenant class. The class must have multi-tenancy enabled. Args: - className: Name of the multi-tenant class - tenants: Array of tenant definitions with name and optional activityStatus Returns the created tenants.
weaviate_update_tenantsUpdate tenant activity status for a multi-tenant class. Use this to activate (HOT/ACTIVE), deactivate (COLD/INACTIVE), freeze (FROZEN), or offload (OFFLOADED) tenants. Args: - className: Name of the multi-tenant class - tenants: Array of tenant updates with name and activityStatus Returns confirmation of update.
weaviate_delete_tenantsDelete tenants from a multi-tenant class. WARNING: This permanently deletes the tenants and all their data. Args: - className: Name of the multi-tenant class - tenantNames: Array of tenant names to delete Returns confirmation of deletion.
weaviate_tenant_existsCheck if a tenant exists in a multi-tenant class. Args: - className: Name of the multi-tenant class - tenantName: Tenant name to check Returns whether the tenant exists.
weaviate_activate_tenantsActivate (warm up) tenants to make them available for queries. This is a convenience method that sets tenant status to ACTIVE. Args: - className: Name of the multi-tenant class - tenantNames: Array of tenant names to activate Returns confirmation of activation.
weaviate_create_backup{
"format": "example_format",
"backend": "example_backend",
"exclude": true,
"include": true,
"backupId": "example_backupId"
}weaviate_get_backup_status{
"format": "example_format",
"backend": "example_backend",
"backupId": "example_backupId"
}weaviate_restore_backup{
"format": "example_format",
"backend": "example_backend",
"exclude": true,
"include": true,
"backupId": "example_backupId"
}