Hive
Industry: Live Streaming
Our Customer
8
Hive Models deployed
in production
Hive Models deployed
in production
12B+
task units processed
in 2025 (excl. VLM)
task units processed
in 2025 (excl. VLM)
25B+
tokens processed
by VLM
tokens processed
by VLM
13M+
posts moderated
in 2025
posts moderated
in 2025
“At Tango, we are incredibly proud of our partnership with Hive for our content moderation needs.
What truly sets this relationship apart is our ability to work closely, side-by-side, to tackle complex challenges. Through this deep collaboration, we are constantly iterating, improving model accuracy, and ensuring that our community remains a safe and welcoming space for everyone.”
What truly sets this relationship apart is our ability to work closely, side-by-side, to tackle complex challenges. Through this deep collaboration, we are constantly iterating, improving model accuracy, and ensuring that our community remains a safe and welcoming space for everyone.”
Dor Isseroff
Chief Operating Officer, Tango
About
About
Tango is one of the world’s largest live streaming platforms, with more than 500 million registered users and over 100 million app downloads.
Tango is one of the world’s largest live streaming platforms, with more than 500 million registered users and over 100 million app downloads.
Challenge
Challenge
As live video became central to the Tango experience, the company faced the challenge of moderating real-time video chats at scale while still fostering a strong sense of community. With users of all ages on the platform, Tango needed more nuanced and granular moderation to better protect younger users and address a broad spectrum of safety risks in live interactions. Real-time comprehensive moderation coverage was essential to preserving user trust.
As live video became central to the Tango experience, the company faced the challenge of moderating real-time video chats at scale while still fostering a strong sense of community. With users of all ages on the platform, Tango needed more nuanced and granular moderation to better protect younger users and address a broad spectrum of safety risks in live interactions. Real-time comprehensive moderation coverage was essential to preserving user trust.
Solution
Solution
To modernize its legacy in-house moderation system, Tango needed a solution built for the speed, scale, and complexity of livestreams. Traditional models could detect certain signals in individual frames, but were less effective at understanding context across a live video feed. Hive’s Vision Language Model (VLM) enabled Tango to analyze multiple frames simultaneously, providing a clearer view of what was happening within a stream and supporting more nuanced moderation decisions. Tango also used Hive’s VLM to improve detection of text violations and streamline policy enforcement through a single API.
In addition, Tango deployed Hive’s Visual Moderation and Text Moderation to strengthen coverage across both video and chat. Visual Moderation provided real-time detection of nudity, violence, and other policy violations in livestream video, helping Tango respond quickly as violative content appeared. Text Moderation extended that coverage to live chat, identifying harassment, solicitation, and harmful language in real time. Together, these systems gave Tango a more scalable approach to livestream safety while helping the team meet evolving safety and compliance requirements.
To modernize its legacy in-house moderation system, Tango needed a solution built for the speed, scale, and complexity of livestreams. Traditional models could detect certain signals in individual frames, but were less effective at understanding context across a live video feed. Hive’s Vision Language Model (VLM) enabled Tango to analyze multiple frames simultaneously, providing a clearer view of what was happening within a stream and supporting more nuanced moderation decisions. Tango also used Hive’s VLM to improve detection of text violations and streamline policy enforcement through a single API.
In addition, Tango deployed Hive’s Visual Moderation and Text Moderation to strengthen coverage across both video and chat. Visual Moderation provided real-time detection of nudity, violence, and other policy violations in livestream video, helping Tango respond quickly as violative content appeared. Text Moderation extended that coverage to live chat, identifying harassment, solicitation, and harmful language in real time. Together, these systems gave Tango a more scalable approach to livestream safety while helping the team meet evolving safety and compliance requirements.
Impact
Impact
Since partnering with Hive, Tango has continuously deployed new AI models and detection capabilities, reducing users’ exposure to harmful content by more than 80%. Each model iteration and implementation has improved detection accuracy, shortened moderation response times, and strengthened the overall safety of the Tango platform.
Since partnering with Hive, Tango has continuously deployed new AI models and detection capabilities, reducing users’ exposure to harmful content by more than 80%. Each model iteration and implementation has improved detection accuracy, shortened moderation response times, and strengthened the overall safety of the Tango platform.
Ready to strengthen content moderation?
Ready to strengthen content moderation?
See how Hive helps platforms detect harmful content, enforce policy, and support trust and safety operations through one comprehensive suite.
See how Hive helps platforms detect harmful content, enforce policy, and support trust and safety operations through one comprehensive suite.