Merge ee823f8488 into 6f24f5a595

docs/docs/configuration/genai/config.md

@@ -9,7 +9,7 @@ import NavPath from "@site/src/components/NavPath";
 
 ## Configuration
 
-A Generative AI provider can be configured in the global config, which will make the Generative AI features available for use. There are currently 4 native providers available to integrate with Frigate. Other providers that support the OpenAI standard API can also be used. See the OpenAI-Compatible section below.
+A Generative AI provider can be configured in the global config, which will make the Generative AI features available for use. There are currently several native providers available to integrate with Frigate. Other providers that support the OpenAI standard API can also be used. See the OpenAI-Compatible section below.
 
 To use Generative AI, you must define a single provider at the global level of your Frigate configuration. If the provider you choose requires an API key, you may either directly paste it in your configuration, or store it in an environment variable prefixed with `FRIGATE_`.
 
@@ -385,3 +385,37 @@ genai:
 
 </TabItem>
 </ConfigTabs>
+
+### TwelveLabs (Marengo embeddings)
+
+[TwelveLabs](https://twelvelabs.io) provides the Marengo multimodal model, which embeds text and images into a shared vector space. This makes it usable as the `embeddings` provider for Frigate's [Semantic Search](/configuration/semantic_search), letting a natural-language query match stored event thumbnails directly.
+
+This provider implements only the `embeddings` role. Use a separate provider (such as Ollama, Gemini, or OpenAI) for the `descriptions` and `chat` roles if you want generated descriptions as well.
+
+#### Get API Key
+
+Create an API key from the [TwelveLabs dashboard](https://playground.twelvelabs.io/dashboard/api-key). There is a generous free tier.
+
+#### Configuration
+
+<ConfigTabs>
+<TabItem value="ui">
+
+1. Navigate to <NavPath path="Settings > Enrichments > Generative AI" />.
+   - Set **Provider** to `twelvelabs`
+   - Set **API key** to your TwelveLabs API key (or use an environment variable such as `{FRIGATE_TWELVELABS_API_KEY}`)
+   - Optionally set **Model** to override the default Marengo model (`marengo3.0`)
+
+</TabItem>
+<TabItem value="yaml">
+
+```yaml
+genai:
+  provider: twelvelabs
+  api_key: "{FRIGATE_TWELVELABS_API_KEY}"
+  roles:
+    - embeddings
+```
+
+</TabItem>
+</ConfigTabs>

frigate/config/camera/genai.py

@@ -15,6 +15,7 @@ class GenAIProviderEnum(str, Enum):
     gemini = "gemini"
     ollama = "ollama"
     llamacpp = "llamacpp"
+    twelvelabs = "twelvelabs"
 
 
 class GenAIRoleEnum(str, Enum):

frigate/genai/plugins/twelvelabs.py

@@ -0,0 +1,139 @@
+"""TwelveLabs provider for Frigate AI.
+
+Provides multimodal embeddings for Frigate's semantic search via TwelveLabs'
+Marengo model. Marengo produces text and image embeddings in a shared vector
+space, so a natural-language query and a stored event thumbnail can be matched
+directly — which is exactly what the ``embeddings`` GenAI role drives.
+
+This provider is opt-in: it is only used when a ``genai`` config entry sets
+``provider: twelvelabs`` and includes the ``embeddings`` role. It does not
+implement the ``descriptions``/``chat`` roles — Marengo is an embedding model,
+and TwelveLabs' Pegasus description model operates on whole video clips rather
+than the per-frame thumbnails Frigate hands to ``_send``.
+
+Marengo embeddings are 512-dimensional; Frigate's semantic search schema
+expects 768 dimensions. The shared :class:`~frigate.embeddings.genai_embedding.GenAIEmbedding`
+adapter zero-pads shorter vectors, so the dimension difference is handled
+upstream and is consistent for both text and image inputs.
+"""
+
+import logging
+
+import numpy as np
+import requests
+
+from frigate.config import GenAIProviderEnum
+from frigate.genai import GenAIClient, register_genai_provider
+
+logger = logging.getLogger(__name__)
+
+# Default Marengo model. Overridable via the `model` config field.
+DEFAULT_MODEL = "marengo3.0"
+
+# Marengo embed REST endpoint. No SDK is needed — this is a plain multipart POST
+# made through Frigate's existing `requests` dependency.
+EMBED_URL = "https://api.twelvelabs.io/v1.3/embed"
+
+
+@register_genai_provider(GenAIProviderEnum.twelvelabs)
+class TwelveLabsClient(GenAIClient):
+    """GenAI client for Frigate using TwelveLabs Marengo embeddings."""
+
+    def _init_provider(self):
+        """Validate config for the TwelveLabs REST provider.
+
+        The provider is just an HTTPS API, so there is no client object to
+        build — the API key is the only thing required. A non-None sentinel is
+        returned so the shared ``ensure_provider``/initialization machinery
+        treats the provider as available.
+        """
+        if not self.genai_config.api_key:
+            logger.error("TwelveLabs provider requires an api_key.")
+            return None
+
+        return self.genai_config.api_key
+
+    @property
+    def _model(self) -> str:
+        return self.genai_config.model or DEFAULT_MODEL
+
+    def list_models(self) -> list[str]:
+        """Marengo is the embedding model exposed by this provider."""
+        return [DEFAULT_MODEL]
+
+    def embed(
+        self,
+        texts: list[str] | None = None,
+        images: list[bytes] | None = None,
+    ) -> list[np.ndarray]:
+        """Generate Marengo embeddings for text and/or images.
+
+        The TwelveLabs embed API embeds a single input per call, so inputs are
+        sent one at a time. Returns one 512-dim float32 vector per input, in
+        order (texts first, then images). The shared GenAIEmbedding adapter
+        pads these to Frigate's 768-dim search schema.
+
+        Calls the Marengo REST endpoint directly via ``requests`` — no SDK.
+        """
+        if self.provider is None:
+            logger.warning(
+                "TwelveLabs provider has not been initialized. Check your configuration."
+            )
+            return []
+
+        results: list[np.ndarray] = []
+
+        for text in texts or []:
+            vector = self._embed_one(text=text)
+            if vector is not None:
+                results.append(vector)
+
+        for image in images or []:
+            vector = self._embed_one(image=image)
+            if vector is not None:
+                results.append(vector)
+
+        return results
+
+    def _embed_one(
+        self, text: str | None = None, image: bytes | None = None
+    ) -> np.ndarray | None:
+        """Embed a single text or image input, returning a float32 vector.
+
+        Posts a multipart form to the Marengo embed endpoint (``model_name`` plus
+        either a ``text`` or an ``image_file`` part). The endpoint requires
+        multipart/form-data, so every field — including text — is passed via
+        ``files`` (the ``(None, value)`` form makes requests emit a multipart
+        text part). ``self.provider`` holds the validated API key. The 512-dim
+        vector is at ``<text|image>_embedding.segments[0].float`` in the JSON
+        response.
+        """
+        headers = {"x-api-key": self.provider}
+        files: dict = {"model_name": (None, self._model)}
+
+        if text is not None:
+            files["text"] = (None, text)
+            result_key = "text_embedding"
+        else:
+            files["image_file"] = ("image.jpg", image, "image/jpeg")
+            result_key = "image_embedding"
+
+        try:
+            response = requests.post(
+                EMBED_URL,
+                headers=headers,
+                files=files,
+                timeout=self.timeout,
+            )
+            response.raise_for_status()
+            result = response.json().get(result_key) or {}
+            segments = result.get("segments") or []
+
+            if not segments:
+                logger.warning("TwelveLabs returned no embedding for input.")
+                return None
+
+            return np.array(segments[0]["float"], dtype=np.float32)
+        except Exception as e:
+            logger.warning("TwelveLabs returned an error: %s", e)
+            return None

frigate/test/test_genai_twelvelabs.py

@@ -0,0 +1,138 @@
+"""Tests for the TwelveLabs GenAI provider (Marengo embeddings)."""
+
+import io
+import os
+import unittest
+from unittest.mock import MagicMock, patch
+
+import numpy as np
+
+from frigate.config.camera.genai import (
+    GenAIConfig,
+    GenAIProviderEnum,
+    GenAIRoleEnum,
+)
+from frigate.genai.plugins.twelvelabs import DEFAULT_MODEL, TwelveLabsClient
+
+
+def _make_config(model: str = "") -> GenAIConfig:
+    return GenAIConfig(
+        provider=GenAIProviderEnum.twelvelabs,
+        api_key="test-key",
+        model=model,
+        roles=[GenAIRoleEnum.embeddings],
+    )
+
+
+def _response(key: str, values):
+    """Mimic the Marengo REST JSON: ``{<key>: {segments: [{float: [...]}]}}``."""
+    resp = MagicMock()
+    resp.raise_for_status.return_value = None
+    resp.json.return_value = {key: {"segments": [{"float": values}]}}
+    return resp
+
+
+class TestTwelveLabsEmbedNoNetwork(unittest.TestCase):
+    """Unit tests with ``requests`` mocked — no network access, no SDK."""
+
+    def _client(self) -> TwelveLabsClient:
+        client = TwelveLabsClient.__new__(TwelveLabsClient)
+        client.genai_config = _make_config()
+        client.timeout = 120
+        client.provider = "test-key"
+        return client
+
+    @patch("frigate.genai.plugins.twelvelabs.requests.post")
+    def test_text_embedding_returns_vector(self, post):
+        post.return_value = _response("text_embedding", [0.1, 0.2, 0.3])
+
+        out = self._client().embed(texts=["a person walking a dog"])
+
+        self.assertEqual(len(out), 1)
+        self.assertIsInstance(out[0], np.ndarray)
+        self.assertEqual(out[0].dtype, np.float32)
+        np.testing.assert_allclose(out[0], [0.1, 0.2, 0.3], rtol=1e-6)
+
+        _, kwargs = post.call_args
+        self.assertEqual(kwargs["files"]["model_name"][1], DEFAULT_MODEL)
+        self.assertEqual(kwargs["files"]["text"][1], "a person walking a dog")
+        self.assertEqual(kwargs["headers"]["x-api-key"], "test-key")
+        self.assertNotIn("image_file", kwargs["files"])
+
+    @patch("frigate.genai.plugins.twelvelabs.requests.post")
+    def test_image_embedding_uses_image_file(self, post):
+        post.return_value = _response("image_embedding", [1.0, 2.0])
+
+        out = self._client().embed(images=[b"\xff\xd8\xff jpeg bytes"])
+
+        self.assertEqual(len(out), 1)
+        _, kwargs = post.call_args
+        self.assertEqual(kwargs["files"]["image_file"][1], b"\xff\xd8\xff jpeg bytes")
+        self.assertNotIn("text", kwargs["files"])
+
+    @patch("frigate.genai.plugins.twelvelabs.requests.post")
+    def test_custom_model_name_is_used(self, post):
+        post.return_value = _response("text_embedding", [0.0])
+
+        client = self._client()
+        client.genai_config = _make_config(model="marengo-custom")
+        client.embed(texts=["x"])
+
+        _, kwargs = post.call_args
+        self.assertEqual(kwargs["files"]["model_name"][1], "marengo-custom")
+
+    @patch("frigate.genai.plugins.twelvelabs.requests.post")
+    def test_empty_segments_are_skipped(self, post):
+        resp = MagicMock()
+        resp.raise_for_status.return_value = None
+        resp.json.return_value = {"text_embedding": {"segments": []}}
+        post.return_value = resp
+
+        self.assertEqual(self._client().embed(texts=["x"]), [])
+
+    @patch("frigate.genai.plugins.twelvelabs.requests.post")
+    def test_api_error_is_swallowed(self, post):
+        post.side_effect = RuntimeError("boom")
+
+        self.assertEqual(self._client().embed(texts=["x"]), [])
+
+    def test_no_provider_returns_empty(self):
+        client = self._client()
+        client.provider = None
+        self.assertEqual(client.embed(texts=["x"]), [])
+
+    def test_no_inputs_returns_empty(self):
+        self.assertEqual(self._client().embed(), [])
+
+
+@unittest.skipUnless(
+    os.environ.get("TWELVELABS_API_KEY"),
+    "TWELVELABS_API_KEY not set; skipping live TwelveLabs API test",
+)
+class TestTwelveLabsEmbedLive(unittest.TestCase):
+    """Live smoke test against the real TwelveLabs API (Marengo)."""
+
+    def _client(self) -> TwelveLabsClient:
+        config = _make_config()
+        config.api_key = os.environ["TWELVELABS_API_KEY"]
+        return TwelveLabsClient(config)
+
+    def test_text_embedding_dim(self):
+        out = self._client().embed(texts=["a delivery person at the front door"])
+        self.assertEqual(len(out), 1)
+        self.assertEqual(out[0].shape, (512,))
+
+    def test_image_embedding_dim(self):
+        from PIL import Image
+
+        arr = (np.random.rand(224, 224, 3) * 255).astype("uint8")
+        buf = io.BytesIO()
+        Image.fromarray(arr, "RGB").save(buf, format="JPEG")
+
+        out = self._client().embed(images=[buf.getvalue()])
+        self.assertEqual(len(out), 1)
+        self.assertEqual(out[0].shape, (512,))
+
+
+if __name__ == "__main__":
+    unittest.main()