feat: voyage-multimodal-3.5 support

docs/integrations/embedding/voyageai.mdx

@@ -3,7 +3,7 @@ title: VoyageAI
 sidebarTitle: VoyageAI
 ---
 
-import { PyEmbeddingVoyageaiUsage } from '/snippets/integrations.mdx';
+import { PyEmbeddingVoyageaiUsage, PyEmbeddingVoyageaiMultimodal } from '/snippets/integrations.mdx';
 
 Voyage AI provides cutting-edge embedding and rerankers.
 
@@ -22,15 +22,20 @@ Supported models are:
 - voyage-multilingual-2
 - voyage-law-2
 - voyage-code-2
+- voyage-multimodal-3.5 (multimodal - supports text, images, and video)
 
+<Info>
+**Multimodal Model:** `voyage-multimodal-3.5` supports text, images, and video inputs. It outputs 1024-dimensional embeddings by default, configurable via the `output_dimension` parameter (256, 512, 1024, 2048). See the [VoyageAI multimodal embeddings documentation](https://docs.voyageai.com/docs/multimodal-embeddings) for more details.
+</Info>
 
 Supported parameters (to be passed in `create` method) are:
 
 | Parameter | Type | Default Value | Description |
 |---|---|--------|---------|
-| `name` | `str` | `None` | The model ID of the model to use. Supported base models for Text Embeddings: voyage-3, voyage-3-lite, voyage-finance-2, voyage-multilingual-2, voyage-law-2, voyage-code-2 |
+| `name` | `str` | `None` | The model ID of the model to use. Supported models: voyage-3, voyage-3-lite, voyage-3.5, voyage-3.5-lite, voyage-context-3, voyage-finance-2, voyage-multilingual-2, voyage-law-2, voyage-code-2, voyage-multimodal-3.5 |
 | `input_type` | `str` | `None` | Type of the input text. Default to None. Other options: query, document. |
 | `truncation` | `bool` | `True` | Whether to truncate the input texts to fit within the context length. |
+| `output_dimension` | `int` | `None` | Output embedding dimension. Only supported by `voyage-multimodal-3.5`. Valid options: 256, 512, 1024 (default), 2048. |
 
 
 Usage Example:
@@ -39,4 +44,14 @@ Usage Example:
   <CodeBlock filename="Python" language="Python" icon="python">
     {PyEmbeddingVoyageaiUsage}
   </CodeBlock>
+</CodeGroup>
+
+### Multimodal Example
+
+The `voyage-multimodal-3.5` model can embed text alongside images. You can use image URLs, file paths, or PIL Image objects:
+
+<CodeGroup>
+  <CodeBlock filename="Python" language="Python" icon="python">
+    {PyEmbeddingVoyageaiMultimodal}
+  </CodeBlock>
 </CodeGroup>

docs/snippets/integrations.mdx

@@ -36,6 +36,8 @@ export const PyEmbeddingOpenclipTextSearch = "actual = table.search(\"man's best
 
 export const PyEmbeddingSentenceTransformersBaai = "import tempfile\nfrom pathlib import Path\n\nimport lancedb\nfrom lancedb.embeddings import get_registry\nfrom lancedb.pydantic import LanceModel, Vector\n\ndb = lancedb.connect(str(Path(tempfile.mkdtemp()) / \"sentence-transformers\"))\nmodel = (\n    get_registry()\n    .get(\"sentence-transformers\")\n    .create(name=\"BAAI/bge-small-en-v1.5\", device=\"cpu\")\n)\n\nclass Words(LanceModel):\n    text: str = model.SourceField()\n    vector: Vector(model.ndims()) = model.VectorField()\n\ntable = db.create_table(\"words\", schema=Words)\ntable.add(\n    [\n        {\"text\": \"hello world\"},\n        {\"text\": \"goodbye world\"},\n    ]\n)\n\nquery = \"greetings\"\nactual = table.search(query).limit(1).to_pydantic(Words)[0]\nprint(actual.text)\n";
 
+export const PyEmbeddingVoyageaiMultimodal = "import tempfile\nfrom pathlib import Path\n\nimport lancedb\nfrom lancedb.embeddings import EmbeddingFunctionRegistry\nfrom lancedb.pydantic import LanceModel, Vector\n\n# Create multimodal embedding function with custom dimension\nvoyageai = (\n    EmbeddingFunctionRegistry.get_instance()\n    .get(\"voyageai\")\n    .create(name=\"voyage-multimodal-3.5\", output_dimension=512)\n)\n\nclass ImageModel(LanceModel):\n    image_uri: str = voyageai.SourceField()\n    vector: Vector(voyageai.ndims()) = voyageai.VectorField()\n\ndb = lancedb.connect(str(Path(tempfile.mkdtemp()) / \"voyageai-multimodal\"))\ntbl = db.create_table(\"images\", schema=ImageModel, mode=\"overwrite\")\n\n# Add images using URLs\ntbl.add(\n    [\n        {\"image_uri\": \"https://upload.wikimedia.org/wikipedia/commons/thumb/4/47/PNG_transparency_demonstration_1.png/300px-PNG_transparency_demonstration_1.png\"},\n    ]\n)\n\n# Search with text query\nresults = tbl.search(\"dice\").limit(1).to_list()\nprint(results)\n";
+
 export const PyEmbeddingVoyageaiUsage = "import tempfile\nfrom pathlib import Path\n\nimport lancedb\nfrom lancedb.embeddings import EmbeddingFunctionRegistry\nfrom lancedb.pydantic import LanceModel, Vector\n\nvoyageai = (\n    EmbeddingFunctionRegistry.get_instance().get(\"voyageai\").create(name=\"voyage-3\")\n)\n\nclass TextModel(LanceModel):\n    text: str = voyageai.SourceField()\n    vector: Vector(voyageai.ndims()) = voyageai.VectorField()\n\ndata = [{\"text\": \"hello world\"}, {\"text\": \"goodbye world\"}]\n\ndb = lancedb.connect(str(Path(tempfile.mkdtemp()) / \"voyageai-demo\"))\ntbl = db.create_table(\"test\", schema=TextModel, mode=\"overwrite\")\n\ntbl.add(data)\n";
 
 export const PyFrameworksLangchainAddImages = "image_uris = [\"./assets/image-1.png\", \"./assets/image-2.png\"]\nvector_store.add_images(uris=image_uris)\n# here image_uris are local fs paths to the images.\n";

tests/py/test_integrations.py

@@ -548,6 +548,44 @@ class TextModel(LanceModel):
     # --8<-- [end:embedding_voyageai_usage]
 
 
+def test_embedding_voyageai_multimodal() -> None:
+    require_env("VOYAGE_API_KEY")
+
+    # --8<-- [start:embedding_voyageai_multimodal]
+    import tempfile
+    from pathlib import Path
+
+    import lancedb
+    from lancedb.embeddings import EmbeddingFunctionRegistry
+    from lancedb.pydantic import LanceModel, Vector
+
+    # Create multimodal embedding function with custom dimension
+    voyageai = (
+        EmbeddingFunctionRegistry.get_instance()
+        .get("voyageai")
+        .create(name="voyage-multimodal-3.5", output_dimension=512)
+    )
+
+    class ImageModel(LanceModel):
+        image_uri: str = voyageai.SourceField()
+        vector: Vector(voyageai.ndims()) = voyageai.VectorField()
+
+    db = lancedb.connect(str(Path(tempfile.mkdtemp()) / "voyageai-multimodal"))
+    tbl = db.create_table("images", schema=ImageModel, mode="overwrite")
+
+    # Add images using URLs
+    tbl.add(
+        [
+            {"image_uri": "https://upload.wikimedia.org/wikipedia/commons/thumb/4/47/PNG_transparency_demonstration_1.png/300px-PNG_transparency_demonstration_1.png"},
+        ]
+    )
+
+    # Search with text query
+    results = tbl.search("dice").limit(1).to_list()
+    print(results)
+    # --8<-- [end:embedding_voyageai_multimodal]
+
+
 # Reranking integrations