T

TwelveLabs

by TwelveLabs, Inc.

AI Models & APIsData & AnalyticsEnterprise Search & KnowledgeVideo & Animation

Video intelligence API that makes every hour of enterprise footage searchable, analyzable and agent-ready.

Freemium · Usage-based · Contact for pricing·Added Jul 14, 2026·Updated Jul 14, 2026
Share:
THE DAILY BRIEF
TwelveLabs

by TwelveLabs, Inc.

AI Models & APIsData & AnalyticsEnterprise Search & KnowledgeVideo & Animation

Video intelligence API that makes every hour of enterprise footage searchable, analyzable and agent-ready.

Freemium · Usage-based · Contact for pricing

TwelveLabs is a video intelligence platform and API that turns raw video into searchable, structured, AI-ready data across vision, audio and language. It is built for media, sports, advertising, security, automotive and government teams that need to search, analyze and reason over large video archives rather than generate new video.

At a Glance

Category
AI Models & APIs
Pricing
Freemium, Usage-based, Contact for pricing
Target Market
CTOs, Heads of AI, Enterprise Developers, Data Scientists, Media & Entertainment Teams, Public Sector
Headquarters
San Francisco, USA
Customers
Enterprise customers across media, sports, advertising, security, automotive and government, including NFL Media, MLSE and Sejong City

Key Features

  • Marengo embedding model
  • Pegasus video-language model
  • Natural-language video search
  • High-speed indexing
  • Search, Embed and Analyze APIs
  • Enterprise fine-tuning

Capabilities

text generation
image generation
video generation
code generation
workflow automation
api access
audio generation
fine tuning
agent orchestration

Use Cases

  • Archive monetization
  • Automated highlights
  • Compliance and evidence review

Ideal For

Best For

  • Natural-language search across large enterprise or broadcast video archives
  • Automated highlight creation and content packaging for sports and media
  • Compliance and brand-safety review of video at scale
  • Evidence management and anomaly detection in security and public-sector footage

Market Analysis

Enterprise-gradeBest-of-breed specialistDeveloper-first

Pros

  • Deep specialization in video understanding, a category most general multimodal APIs handle shallowly
  • Transparent usage-based pricing and a real free tier make evaluation easy
  • Strategic AWS partnership and strong media/sports logo list

Cons

  • Narrow to video — not a general-purpose AI platform
  • Cloud-only; no on-premise option published, which can be a blocker for sensitive footage
  • Accuracy figures such as the 78.5% composite score are vendor-reported

Pricing

Free

$0

  • 600 minutes of video indexing
  • No credit card required
  • Search, Embed and Analyze APIs
  • Up to 100 videos, 5 concurrent indexing tasks
  • Index access for 90 days

Developer

From $0.042/min

  • Pay-as-you-go
  • Video indexing $0.042/minute plus $0.0015/minute infrastructure fee
  • Search API $4 per 1,000 queries
  • Analyze API $0.0292/minute of input video
  • Unlimited video hours, 100,000 videos per index, 25 concurrent tasks

Enterprise

Contact for pricing

  • Committed contracts with negotiable rates
  • Model fine-tuning
  • Enterprise support

Free tier includes 600 minutes of indexing with no credit card. Developer plan is pure usage-based; Enterprise rates are negotiated.

THE DAILY BRIEF

Enterprise AI insights for technology and business leaders, twice weekly.

beri.net

Subscribe at beri.net/subscribe for twice-weekly AI insights delivered to your inbox.

LinkedIn: linkedin.com/in/rberi  |  X: x.com/rajeshberi

© 2026 Rajesh Beri. All rights reserved.

TwelveLabs is a video intelligence platform and API that turns raw video into searchable, structured, AI-ready data across vision, audio and language. It is built for media, sports, advertising, security, automotive and government teams that need to search, analyze and reason over large video archives rather than generate new video.

TwelveLabs builds a video intelligence platform and API for enterprises sitting on large video archives. Its two foundation models do the work: Marengo, a multimodal video embedding model that converts video into spatiotemporal embeddings for precise moment-level search (the company reports 78.5% composite accuracy across 47 languages, with Marengo 3.0 the current generation), and Pegasus, a video-language model that reasons continuously over full temporal sequences of up to two hours and converts video into structured data — scene boundaries, entities, temporal segments and semantic context. The result is natural-language search across an entire video library, content segmentation, compliance scanning and archive analysis, at roughly 60x real-time indexing speed ('index an hour of video in a minute'). Notably, TwelveLabs positions video understanding as a distinct category from video generation. The San Francisco and Seoul-based company raised a $100M Series B announced July 1, 2026, co-led by NEA and NAVER Ventures with participation from Amazon, Radical Ventures, Korea Investment Partners, Index Ventures, Quadrille Capital and Red Bull Ventures — bringing total funding to roughly $150M — alongside a strategic partnership making AWS its preferred cloud provider, and is opening offices in New York and London. Named customers and partners include NFL Media, MLSE, Sejong City, MindsDB, Voxel51 and Source Digital. It offers a free tier (600 minutes of indexing, no credit card), a pay-as-you-go Developer plan (video indexing from $0.042/minute; Search API $4 per 1,000 queries) and a custom Enterprise plan with model fine-tuning.

At a Glance

Category
AI Models & APIs
Pricing
Freemium, Usage-based, Contact for pricing
Target Market
CTOs, Heads of AI, Enterprise Developers, Data Scientists, Media & Entertainment Teams, Public Sector
Headquarters
San Francisco, USA
Customers
Enterprise customers across media, sports, advertising, security, automotive and government, including NFL Media, MLSE and Sejong City

Key Features

  • Marengo embedding model

    A multimodal video embedding model that converts vision, audio and text into spatiotemporal embeddings for moment-level retrieval, reported at 78.5% composite accuracy across 47 languages.

  • Pegasus video-language model

    Reasons continuously over video sequences up to two hours long and outputs structured data such as scene boundaries, entities and temporal segments.

  • Natural-language video search

    Query an entire video library in plain language and jump to the exact moment, rather than relying on manual tags or metadata.

  • High-speed indexing

    Indexes video at roughly 60x real time — about one hour of footage per minute.

  • Search, Embed and Analyze APIs

    Three core APIs let developers embed video retrieval, similarity and analysis into their own applications and agents.

  • Enterprise fine-tuning

    The Enterprise plan supports model fine-tuning on a customer's own footage and taxonomy.

Capabilities

text generation
image generation
video generation
code generation
workflow automation
api access
audio generation
fine tuning
agent orchestration

Use Cases

  • Archive monetization

    Make decades of unsearchable footage discoverable so it can be licensed, repackaged and resold.

  • Automated highlights

    Sports and media teams (NFL Media, MLSE) find and cut key moments from live and archived footage without manual logging.

  • Compliance and evidence review

    Scan video for policy violations, anomalies or specific events for compliance, security and public-sector evidence workflows.

Ideal For

Best For

  • Natural-language search across large enterprise or broadcast video archives
  • Automated highlight creation and content packaging for sports and media
  • Compliance and brand-safety review of video at scale
  • Evidence management and anomaly detection in security and public-sector footage

Integrations

SDK Available
SDK:Python

Market & Ratings

Estimated Customers

Enterprise customers across media, sports, advertising, security, automotive and government, including NFL Media, MLSE and Sejong City

Market Analysis

Enterprise-gradeBest-of-breed specialistDeveloper-first

Pros

  • Deep specialization in video understanding, a category most general multimodal APIs handle shallowly
  • Transparent usage-based pricing and a real free tier make evaluation easy
  • Strategic AWS partnership and strong media/sports logo list

Cons

  • Narrow to video — not a general-purpose AI platform
  • Cloud-only; no on-premise option published, which can be a blocker for sensitive footage
  • Accuracy figures such as the 78.5% composite score are vendor-reported

Pricing

Free Trial Available

Free

$0

  • 600 minutes of video indexing
  • No credit card required
  • Search, Embed and Analyze APIs
  • Up to 100 videos, 5 concurrent indexing tasks
  • Index access for 90 days

Developer

From $0.042/min

  • Pay-as-you-go
  • Video indexing $0.042/minute plus $0.0015/minute infrastructure fee
  • Search API $4 per 1,000 queries
  • Analyze API $0.0292/minute of input video
  • Unlimited video hours, 100,000 videos per index, 25 concurrent tasks

Enterprise

Contact for pricing

  • Committed contracts with negotiable rates
  • Model fine-tuning
  • Enterprise support

Free tier includes 600 minutes of indexing with no credit card. Developer plan is pure usage-based; Enterprise rates are negotiated.

Sources

This page was written from 3 sources, 1 on domains other than twelvelabs.io.

  1. 1.twelvelabs.iotwelvelabs.iovendor
  2. 2.twelvelabs.iopricingvendor
  3. 3.globenewswire.comtwelvelabs raises 100 million in series b funding to build v
Newsletter

Stay Ahead of the Curve

Weekly enterprise AI insights for technology leaders. No spam, no vendor pitches—unsubscribe anytime.

Subscribe