Jina

https://jina.ai/

Maven Dependency

<dependency>
    <groupId>dev.langchain4j</groupId>
    <artifactId>langchain4j-jina</artifactId>
    <version>1.19.0-beta29</version>
</dependency>

APIs

• JinaEmbeddingModel

Capabilities

• Multimodal (jina-clip-v2, jina-embeddings-v4 — auto-detected from the model name): embeds text and images. Jina embeds one modality per input item (it does not fuse interleaved text + image); pass a single TextContent or a single ImageContent (URL or base64) per input in an EmbeddingRequest.
• Input type (jina-embeddings-v3, jina-embeddings-v4, jina-embeddings-v5 — auto-detected from the model name): embeds a search query differently from the documents it is matched against, which usually improves retrieval quality. Set EmbeddingInputType.QUERY or DOCUMENT per call on an EmbeddingRequest, or via EmbeddingStoreContentRetriever.embeddingInputType(...). Models that do not offer this (for example jina-clip-v2) reject the parameter instead of silently ignoring it.
• Listeners: configure via JinaEmbeddingModel.builder().listeners(...).

See Embedding Model for the request/response API and multimodal usage.

Examples

• JinaEmbeddingModelIT