# Multimodal (CLIP) Embeddings Weaviate's integration with the Hugging Face Transformers library allows you to access their CLIP models' capabilities directly from Weaviate. [Configure a Weaviate vector index](#configure-the-vectorizer) to use the CLIP integration, and [configure the Weaviate instance](#weaviate-configuration) with a model image, and Weaviate will generate embeddings for various operations using the specified model in the CLIP inference container. This feature is called the _vectorizer_. At [import time](#data-import), Weaviate generates multimodal object embeddings and saves them into the index. For [vector](#vector-near-text-search) and [hybrid](#hybrid-search) search operations, Weaviate converts queries of one or more modalities into embeddings. [Multimodal search operations](#vector-near-media-search) are also supported. ![Embedding integration illustration](/assets/docs/weaviate/model-providers/_includes/integration_transformers_embedding.png) ## Requirements ### Weaviate configuration Your Weaviate instance must be configured with the CLIP multimodal vectorizer integration (`multi2vec-clip`) module. :::accordion{title="For Weaviate Cloud (WCD) users"} This integration is not available for Weaviate Cloud (WCD) instances, as it requires spinning up a container with the Hugging Face model. ::: #### Enable the integration module - Check the [cluster metadata](../monitoring-and-logging/status.md#cluster-metadata) to verify if the module is enabled. - Follow the [how-to configure modules](../how-to-configure-weaviate/modules.md) guide to enable the module in Weaviate. #### Configure the integration To use this integration, configure the container image of the CLIP model and the inference endpoint of the containerized model. The following example shows how to configure the CLIP integration in Weaviate: ::::tabs{sync="deployments"} :::tab{title="Docker"} #### Docker Option 1: Use a pre-configured `docker-compose.yml` file Follow the instructions on the [Weaviate Docker installation configurator](../installation/installation-guides-docker-installation.md#configurator) to download a pre-configured `docker-compose.yml` file with a selected model #### Docker Option 2: Add the configuration manually Alternatively, add the configuration to the `docker-compose.yml` file manually as in the example below. ```yaml services: weaviate: # Other Weaviate configuration environment: CLIP_INFERENCE_API: http://multi2vec-clip:8080 # Set the inference API endpoint multi2vec-clip: # Set the name of the inference container image: cr.weaviate.io/semitechnologies/multi2vec-clip:sentence-transformers-clip-ViT-B-32-multilingual-v1 environment: ENABLE_CUDA: 0 # Set to 1 to enable ``` - `CLIP_INFERENCE_API` environment variable sets the inference API endpoint - `multi2vec-clip` is the name of the inference container - `image` is the container image - `ENABLE_CUDA` environment variable enables GPU usage Set `image` from a [list of available models](#available-models) to specify a particular model to be used. ::: :::tab{title="Kubernetes"} Configure the Hugging Face Transformers integration in Weaviate by adding or updating the `multi2vec-clip` module in the `modules` section of the Weaviate Helm chart values file. For example, modify the `values.yaml` file as follows: ```yaml modules: multi2vec-clip: enabled: true tag: sentence-transformers-clip-ViT-B-32-multilingual-v1 repo: semitechnologies/multi2vec-clip registry: cr.weaviate.io envconfig: enable_cuda: true ``` See the [Weaviate Helm chart](https://github.com/weaviate/weaviate-helm/blob/master/weaviate/values.yaml) for an example of the `values.yaml` file including more configuration options. Set `tag` from a [list of available models](#available-models) to specify a particular model to be used. ::: :::: ### Credentials As this integration runs a local container with the CLIP model, no additional credentials (e.g. API key) are required. Connect to Weaviate as usual, such as in the examples below. :::code-group{sync="languages"} ```python title="Python" ``` ```typescript title="JavaScript/TypeScript" ``` ::: ## Configure the vectorizer [Configure a Weaviate index](../how-to-manage-collections/vector-config.md#specify-a-vectorizer) as follows to use a CLIP embedding model: :::code-group{sync="languages"} ```python title="Python" {9-20} from weaviate.classes.config import Configure, DataType, Multi2VecField, Property client.collections.create( "DemoCollection", properties=[ Property(name="title", data_type=DataType.TEXT), Property(name="poster", data_type=DataType.BLOB), ], vector_config=[ Configure.Vectors.multi2vec_clip( name="title_vector", # Define the fields to be used for the vectorization - using image_fields, text_fields, video_fields image_fields=[ Multi2VecField(name="poster", weight=0.9) ], text_fields=[ Multi2VecField(name="title", weight=0.1) ] ) ], # Additional parameters not shown ) ``` ```typescript title="JavaScript/TypeScript" {13-29} await client.collections.create({ name: 'DemoCollection', properties: [ { name: 'title', dataType: 'text' as const, }, { name: 'poster', dataType: 'blob' as const, }, ], vectorizers: [ weaviate.configure.vectors.multi2VecClip({ name: 'title_vector', imageFields: [ { name: 'poster', weight: 0.9, }, ], textFields: [ { name: 'title', weight: 0.1, }, ], }), ], // Additional parameters not shown }); ``` ::: :::callout{intent="note" title="Choose a container image to select a model"} To choose a model, select the [container image](#configure-the-integration) that hosts it. ::: :::accordion{title="Vectorization behavior"} Weaviate follows the collection configuration and a set of predetermined rules to vectorize objects. Unless specified otherwise in the collection definition, the default behavior is to: - Only vectorize properties that use the `text` or `text[]` data type (unless [skipped](../how-to-manage-collections/vector-config.md#property-level-settings)) - Sort properties in alphabetical (a-z) order before concatenating values - If `vectorizePropertyName` is `true` (`false` by default) prepend the property name to each property value - Join the (prepended) property values with spaces - Prepend the class name (unless `vectorizeClassName` is `false`) - Convert the produced string to lowercase ::: ### Vectorizer parameters #### Inference URL parameters Optionally, if your stack includes multiple inference containers, specify the inference container(s) to use with a collection. If no parameters are specified, the default inference URL from the Weaviate configuration is used. Specify `inferenceUrl` for a single inference container. :::code-group{sync="languages"} ```python title="Python" {9-21} from weaviate.classes.config import Configure, DataType, Multi2VecField, Property client.collections.create( "DemoCollection", properties=[ Property(name="title", data_type=DataType.TEXT), Property(name="poster", data_type=DataType.BLOB), ], vector_config=[ Configure.Vectors.multi2vec_clip( name="title_vector", # Define the fields to be used for the vectorization - using image_fields, text_fields, video_fields image_fields=[ Multi2VecField(name="poster", weight=0.9) ], text_fields=[ Multi2VecField(name="title", weight=0.1) ], # inference_url="" ) ], # Additional parameters not shown ) ``` ```typescript title="JavaScript/TypeScript" {13-30} await client.collections.create({ name: 'DemoCollection', properties: [ { name: 'title', dataType: 'text' as const, }, { name: 'poster', dataType: 'blob' as const, }, ], vectorizers: [ weaviate.configure.vectors.multi2VecClip({ name: 'title_vector', imageFields: [ { name: 'poster', weight: 0.9, }, ], textFields: [ { name: 'title', weight: 0.1, }, ], // inferenceUrl: '' }), ], }); ``` ::: ## Data import After configuring the vectorizer, [import data](../how-to-manage-objects/import.md) into Weaviate. Weaviate generates embeddings for the objects using the specified model. :::code-group{sync="languages"} ```python title="Python" {11-15} collection = client.collections.use("DemoCollection") with collection.batch.fixed_size(batch_size=200) as batch: for src_obj in source_objects: poster_b64 = url_to_base64(src_obj["poster_path"]) weaviate_obj = { "title": src_obj["title"], "poster": poster_b64 # Add the image in base64 encoding } # The model provider integration will automatically vectorize the object batch.add_object( properties=weaviate_obj, # vector=vector # Optionally provide a pre-obtained vector ) ``` ```typescript title="JavaScript/TypeScript" const collectionName = 'DemoCollection' const myCollection = client.collections.use(collectionName) ``` ::: :::callout{intent="tip" title="Re-use existing vectors"} If you already have a compatible model vector available, you can provide it directly to Weaviate. This can be useful if you have already generated embeddings using the same model and want to use them in Weaviate, such as when migrating data from another system. ::: ## Searches Once the vectorizer is configured, Weaviate will perform vector and hybrid search operations using the specified CLIP model. ![Embedding integration at search illustration](/assets/docs/weaviate/model-providers/_includes/integration_transformers_embedding_search.png) ### Vector (near text) search When you perform a [vector search](../how-to-query-search/similarity.md#search-with-text), Weaviate converts the text query into an embedding using the specified model and returns the most similar objects from the database. The query below returns the `n` most similar objects from the database, set by `limit`. :::code-group{sync="languages"} ```python title="Python" {3-6} collection = client.collections.use("DemoCollection") response = collection.query.near_text( query="A holiday film", # The model provider integration will automatically vectorize the query limit=2 ) for obj in response.objects: print(obj.properties["title"]) ``` ```typescript title="JavaScript/TypeScript" const collectionName = 'DemoCollection' const myCollection = client.collections.use(collectionName) ``` ::: ### Hybrid search :::callout{intent="info" title="What is a hybrid search?"} A hybrid search performs a vector search and a keyword (BM25) search, before [combining the results](../how-to-query-search/hybrid.md) to return the best matching objects from the database. ::: When you perform a [hybrid search](../how-to-query-search/hybrid.md), Weaviate converts the text query into an embedding using the specified model and returns the best scoring objects from the database. The query below returns the `n` best scoring objects from the database, set by `limit`. :::code-group{sync="languages"} ```python title="Python" {3-6} collection = client.collections.use("DemoCollection") response = collection.query.hybrid( query="A holiday film", # The model provider integration will automatically vectorize the query limit=2 ) for obj in response.objects: print(obj.properties["title"]) ``` ```typescript title="JavaScript/TypeScript" const collectionName = 'DemoCollection' const myCollection = client.collections.use(collectionName) ``` ::: ### Vector (near media) search When you perform a media search such as a [near image search](../how-to-query-search/similarity.md#search-with-image), Weaviate converts the query into an embedding using the specified model and returns the most similar objects from the database. To perform a near media search such as near image search, convert the media query into a base64 string and pass it to the search query. The query below returns the `n` most similar objects to the input image from the database, set by `limit`. :::code-group{sync="languages"} ```python title="Python" def url_to_base64(url): import requests import base64 image_response = requests.get(url) content = image_response.content return base64.b64encode(content).decode("utf-8") ``` ```typescript title="JavaScript/TypeScript" const base64String = 'SOME_BASE_64_REPRESENTATION'; result = await myCollection.query.nearImage( base64String, // The model provider integration will automatically vectorize the query { limit: 2, } ) console.log(JSON.stringify(result.objects, null, 2)); ``` ::: ## References ### Available models Lists of pre-built Docker images for this integration are below. | Model Name | Image Name | Notes | | --------------------------------------------------- | ---------------------------------------------------------------------------------------------------- | ------------------------------------------------------------------------------------------------------------------------------------------- | | google/siglip2-so400m-patch16-512 | `cr.weaviate.io/semitechnologies/multi2vec-clip:google-siglip2-so400m-patch16-512` | SigLIP 2 model with 512x512 input size, added in `multi2vec-clip` `v1.4.0` (Multilingual, 1152d) | | google/siglip2-so400m-patch16-384 | `cr.weaviate.io/semitechnologies/multi2vec-clip:google-siglip2-so400m-patch16-384` | SigLIP 2 model with 384x384 input size, added in `multi2vec-clip` `v1.4.0` (Multilingual, 1152d) | | sentence-transformers-clip-ViT-B-32 | `cr.weaviate.io/semitechnologies/multi2vec-clip:sentence-transformers-clip-ViT-B-32` | Texts must be in English. (English, 768d) | | sentence-transformers-clip-ViT-B-32-multilingual-v1 | `cr.weaviate.io/semitechnologies/multi2vec-clip:sentence-transformers-clip-ViT-B-32-multilingual-v1` | Supports a wide variety of languages for text. See sbert.net for details. (Multilingual, 768d) | | openai-clip-vit-base-patch16 | `cr.weaviate.io/semitechnologies/multi2vec-clip:openai-clip-vit-base-patch16` | The base model uses a ViT-B/16 Transformer architecture as an image encoder and uses a masked self-attention Transformer as a text encoder. | | ViT-B-16-laion2b\_s34b\_b88k | `cr.weaviate.io/semitechnologies/multi2vec-clip:ViT-B-16-laion2b_s34b_b88k` | The base model uses a ViT-B/16 Transformer architecture as an image encoder trained with LAION-2B dataset using OpenCLIP. | | ViT-B-32-quickgelu-laion400m\_e32 | `cr.weaviate.io/semitechnologies/multi2vec-clip:ViT-B-32-quickgelu-laion400m_e32` | The base model uses a ViT-B/32 Transformer architecture as an image encoder trained with LAION-400M dataset using OpenCLIP. | | xlm-roberta-base-ViT-B-32-laion5b\_s13b\_b90k | `cr.weaviate.io/semitechnologies/multi2vec-clip:xlm-roberta-base-ViT-B-32-laion5b_s13b_b90k` | Uses ViT-B/32 xlm roberta base model trained with the LAION-5B dataset using OpenCLIP. | We add new model support over time. For a complete list of available models, see the Docker Hub tags for the [multi2vec-clip](https://hub.docker.com/r/semitechnologies/multi2vec-clip/tags) container. ## Advanced configuration ### Run a separate inference container As an alternative, you can run the inference container independently from Weaviate. To do so, follow these steps: - Enable `multi2vec-clip` and omit `multi2vec-clip` container parameters in your [Weaviate configuration](#weaviate-configuration) - Run the inference container separately, e.g. using Docker, and - Use `CLIP_INFERENCE_API` or [`inferenceUrl`](#configure-the-vectorizer) to set the URL of the inference container. For example, run the container with Docker: ```shell docker run -itp "8000:8080" semitechnologies/multi2vec-clip:sentence-transformers-clip-ViT-B-32-multilingual-v1 ``` Then, set `CLIP_INFERENCE_API="http://localhost:8000"`. If Weaviate is part of the same Docker network, as a part of the same `docker-compose.yml` file, you can use the Docker networking/DNS, such as `CLIP_INFERENCE_API=http://multi2vec-clip:8080`. ## Further resources ### Other integrations - [Transformers text embedding models + Weaviate](transformers-embeddings.md). - [Transformers reranker models + Weaviate](transformers-reranker.md). ### Code examples Once the integrations are configured at the collection, the data management and search operations in Weaviate work identically to any other collection. See the following model-agnostic examples: - The [How-to: Manage collections](../how-to-manage-collections/index.md) and [How-to: Manage objects](../how-to-manage-objects/index.md) guides show how to perform data operations (i.e. create, read, update, delete collections and objects within them). - The [How-to: Query & Search](../how-to-query-search/index.md) guides show how to perform search operations (i.e. vector, keyword, hybrid) as well as retrieval augmented generation. ### Model licenses Each of the compatible models has its own license. For detailed information, review the license for the model you are using in the [Hugging Face Model Hub](https://huggingface.co/models). It is your responsibility to evaluate whether the terms of its license(s), if any, are appropriate for your intended use. ### External resources - Hugging Face [Model Hub](https://huggingface.co/models) ## Questions and feedback Have a question or feedback? Here's how to reach us. ::::card-grid :::card{title="Community Forum" href="https://forum.weaviate.io/c/support" icon="messages-square"} Ask questions and connect with other developers on our **Community forum**. ::: :::card{title="Support" href="/guides/support-overview" icon="life-buoy"} Weaviate Cloud user or customer? Find the right channel on the **Support page**. ::: :::: ## Related pages - [Agents](./agents-index.md) - [AI-assisted Weaviate code generation](./ai-assisted-vibe-coding-index.md) - [APIs](./apis-index.md) - [Authorization and authentication](./authorization-and-authentication-index.md) - [Benchmarks](./benchmarks-index.md) - [Best practices](./best-practices-index.md) - [Client libraries](./clients-index.md) - [Client Libraries / SDKs](./client-libraries-index.md) - [Cloud](./cloud-index.md) - [Cloud account management](./cloud-account-management-index.md) # Agent Instructions This portal answers questions programmatically. To receive a synthesized, source-cited answer instead of crawling page by page, append the `?ask=` query parameter to any page URL on this site: /guides/quickstart?ask=how+do+I+authenticate Optional parameters: - `&goal=` steers the answer toward your objective (e.g. `&goal=write+a+python+client`). - `&version=