Milvus

Configure your own Milvus Vector DB instance as data sink.

Configure via UI

Configure via API / Client

```python from llama_cloud.types import CloudMilvusVectorStore

ds = { ‘name’: '', ‘sink_type’: ‘MILVUS’, ‘component’: CloudMilvusVectorStore( uri='', collection_name=‘<collection_name>’, token='', # optional # embedding dimension dim='' # optional ) } data_sink = client.data_sinks.create_data_sink(request=ds)

</TabItem>
<TabItem value="typescript" label="TypeScript Client" default>
```Typescript
const ds = {
  'name': 'milvus',
  'sinkType': 'MILVUS',
  'component': {
      'uri': '<uri>',
      'collection_name': '<collection_name>',
      'token': '<token>',  // optional
      // embedding dimension
      'dim': '<dim>'  // optional
  }
}

data_sink = await client.dataSinks.createDataSink({
projectId: projectId,
body: ds
})