TwelveLabs | Embed

Embed

NEW

Vectorize everything.
Do anything.

Turn rich video data – including image, text, and audio – into vectors and new possibilities. Easily build semantic search, hybrid search, recommender systems, anomaly detection, and more.

Try on Playground

🎉 TwelveLabs models are now in Amazon Bedrock! Leading video understanding meets with the scale of AWS infrastructure. Learn more here!

Products

Enterprise

Research

Developers

Company

Select Language

Playground

Talk to Sales

🎉 TwelveLabs models are now in Amazon Bedrock! Leading video understanding meets with the scale of AWS infrastructure. Learn more here!

Embed 01

Multimodal doesn’t have to mean multi-model.

No more piecing together siloed solutions for image, text, audio and video. Support all modalities and turn rich video data into vectors in the same space.

Embed 01

Multimodal doesn’t have to mean multi-model.

No more piecing together siloed solutions for image, text, audio and video. Support all modalities and turn rich video data into vectors in the same space.

Embed 02

Simple doesn’t have to mean generic.

Your data is unique – your models should be, too. Fine-tune our models easily for your domain until they deliver unparalleled performance.

Embed 02

Simple doesn’t have to mean generic.

Your data is unique – your models should be, too. Fine-tune our models easily for your domain until they deliver unparalleled performance.

Embed 03

Better output with shorter processing times.

With native video support, Embed API reduces processing time, increasing throughput, and saving you time and money.

Embed 03

Better output with shorter processing times.

With native video support, Embed API reduces processing time, increasing throughput, and saving you time and money.

For everything your video can do.

RAG pairing

Pair our models with your RAG pipeline to retrieve relevant information and improve data output.

High-quality training data

Transform workflows with embeddings to create training data, improve data quality, and reduce manual labeling needs.

Training models

Use embeddings to improve data quality when training large language models.

Anomaly detection

Codentify anomalies – for example, detect and remove corrupt videos that only display a black background – to enhance data quality.

For everything your video can do.

RAG pairing

Pair our models with your RAG pipeline to retrieve relevant information and improve data output.

High-quality training data

Transform workflows with embeddings to create training data, improve data quality, and reduce manual labeling needs.

Customer Search

Let customers easily find   any video moment within   your platform.

Asset Management

Comb through petabytes   of data using natural   language queries.

For everything your video can do.

RAG pairing

Pair our models with your RAG pipeline to retrieve relevant information and improve data output.

High-quality training data

Transform workflows with embeddings to create training data, improve data quality, and reduce manual labeling needs.

Customer Search

Let customers easily find   any video moment within   your platform.

Asset Management

Comb through petabytes   of data using natural   language queries.

Sample Apps

node

PYTHON

Contextual and Personalized Ads

A tool for analyzing source footage, summarizing content, and recommending ads based on the footage's context and emotional tone.

Try this sample app

PYTHON

Recommendations using Multimodal Embeddings

Start exploring videos and discovering similar content powered by TwelveLabs Multimodal Embeddings.

Try this sample app

Python

Node

from twelvelabs import TwelveLabs
from twelvelabs.models.embed import EmbeddingsTask, SegmentEmbedding

client = TwelveLabs("<YOUR_API_KEY>")

# Create a video embedding task for your video
task = client.embed.task.create(
   model_name="Marengo-retrieval-2.7",
   video_url: "<YOUR_VIDEO_URL>"
)

print(f"Created task: id={task.id} model_name={task.model_name} status={task.status}")

# Wait for embedding task to finish
status = task.wait_for_done()
print(f"Embedding done: {status}")

# Retrieve the video embeddings
task = task.retrieve()

# Print the embeddings
if task.video_embedding is not None and task.video_embedding.segments is not None:
    for segment in task.video_embedding.segments:
            print(
                f"  embedding_scope={segment.embedding_scope} start_offset_sec={segment.start_offset_sec} end_offset_sec={segment.end_offset_sec}"
            )
            print(f"  embeddings: {", ".join(str(segment.embeddings_float))}")

Python

Node

from twelvelabs import TwelveLabs
from twelvelabs.models.embed import EmbeddingsTask, SegmentEmbedding

client = TwelveLabs("<YOUR_API_KEY>")

# Create a video embedding task for your video
task = client.embed.task.create(
   model_name="Marengo-retrieval-2.7",
   video_url: "<YOUR_VIDEO_URL>"
)

print(f"Created task: id={task.id} model_name={task.model_name} status={task.status}")

# Wait for embedding task to finish
status = task.wait_for_done()
print(f"Embedding done: {status}")

# Retrieve the video embeddings
task = task.retrieve()

# Print the embeddings
if task.video_embedding is not None and task.video_embedding.segments is not None:
    for segment in task.video_embedding.segments:
            print(
                f"  embedding_scope={segment.embedding_scope} start_offset_sec={segment.start_offset_sec} end_offset_sec={segment.end_offset_sec}"
            )
            print(f"  embeddings: {", ".join(str(segment.embeddings_float))}")

Python

Node

from twelvelabs import TwelveLabs
from twelvelabs.models.embed import EmbeddingsTask, SegmentEmbedding

client = TwelveLabs("<YOUR_API_KEY>")

# Create a video embedding task for your video
task = client.embed.task.create(
   model_name="Marengo-retrieval-2.7",
   video_url: "<YOUR_VIDEO_URL>"
)

print(f"Created task: id={task.id} model_name={task.model_name} status={task.status}")

# Wait for embedding task to finish
status = task.wait_for_done()
print(f"Embedding done: {status}")

# Retrieve the video embeddings
task = task.retrieve()

# Print the embeddings
if task.video_embedding is not None and task.video_embedding.segments is not None:
    for segment in task.video_embedding.segments:
            print(
                f"  embedding_scope={segment.embedding_scope} start_offset_sec={segment.start_offset_sec} end_offset_sec={segment.end_offset_sec}"
            )
            print(f"  embeddings: {", ".join(str(segment.embeddings_float))}")

Integrate with your personalized SDK — and your vision.

Deploy your custom-trained model on any cloud. See and surface everything in your video, then go beyond with AI that can realize your most game-changing ideas.

Get the SDK