Young Zhao is a technology entrepreneur and the co-founder and CEO of OpusClip, an AI-powered video clipping and editing company headquartered in Mountain View, California.
He is one of three publicly identified founders of OpusClip, alongside Jay Wu and Grace Wang. OpusClip's own public materials and investor records identify the company as founded in 2022, while OpusClip states that the standalone OpusClip clipping product launched in June 2023.
Zhao's role is particularly important to understanding the origin of OpusClip because his public interviews provide a detailed account of the company's original problem, its first product, its early failure, and the product pivot that eventually became OpusClip.
Importantly, however, the available evidence does not support describing Zhao as the sole inventor of OpusClip or as the person who personally developed every underlying AI model. OpusClip was founded by a three-person team and subsequently developed by a larger engineering and research organization.
Education and Early Career
Public professional records identify Young Zhao as a graduate of the University of California, Berkeley, where he studied Computer Science with a focus on Human-Computer Interaction.
Before OpusClip, Zhao was involved in entrepreneurship and technology-oriented product development. Public professional records associate him with Zaihui and OneBowl, while other accounts describe his experience building and operating a social-media talent agency.
That operating experience became central to the creation of OpusClip.
Zhao has explained that while running his agency, he observed the growing number of people who wanted to become independent content creators. Producing video for clients was already labor-intensive, but the rise of creators producing content for their own audiences created another problem: there was enormous demand for video, while editing remained expensive, slow and difficult.
This experience led him toward the idea of using artificial intelligence to automate parts of the content-production process.
The Original OpusClip Idea
OpusClip did not begin as exactly the product that exists today.
According to Zhao's account of the company's early development, the team founded the company in January 2022 and initially pursued an AI-powered livestreaming product.
The original product included AI-based interactive capabilities such as automatically generated memes and other livestream interactions. The team spent roughly six months building the product.
The initial product, however, failed to demonstrate strong product-market fit.
Zhao has described how the livestreaming product attracted only around 200 users during its early period. The team attempted to collect feedback from those users, including paying some participants to use the product and provide feedback.
The most important discovery came from a feature that was not originally intended to be the entire business.
Users wanted to reuse their livestream recordings after the livestream ended. They wanted the system to identify useful moments and turn them into clips that could be distributed through other channels.
The clipping functionality became the part of the product that users actually valued.
The team therefore made a major strategic decision: instead of continuing to build the livestreaming product, it would turn the clipping functionality into a standalone product.
That pivot became OpusClip.
From Livestreaming Tool to OpusClip
The transformation from the original livestreaming product into OpusClip is one of the most important events in the company's history.
Zhao has said that once the clipping functionality became a standalone product, tens of thousands of users signed up within approximately two weeks.
OpusClip subsequently launched publicly in June 2023.
The company's own December 2023 announcement stated that OpusClip had already reached more than one million creators and businesses only months after launch.
By the company's second anniversary in June 2025, OpusClip reported that more than 12 million creators and brands had used the platform.
The company's current website reports more than 20 million creators and businesses and more than 440 million clips created, with those figures marked as of September 2026.
The evolution illustrates an important distinction in startup history: the company was founded in 2022, but the product that became the core OpusClip business emerged through experimentation and a subsequent pivot from the original livestreaming concept.
Who Founded OpusClip?
The publicly documented founding team consists of:
- Young Zhao — Co-Founder and CEO
- Jay Wu — Co-Founder and CTO
- Grace Wang — Co-Founder and CMO
ZhenFund's official portfolio independently lists OpusClip as founded in 2022 and identifies Young Zhao, Grace Wang and Jay W as its founders.
The three founders brought different backgrounds to the company.
Zhao's background combines computer science, human-computer interaction, entrepreneurship and creator-economy experience.
Jay Wu's documented background is substantially more technical. Public professional records identify him as OpusClip's co-founder and CTO and list a bachelor's degree in Electrical and Computer Engineering and Computer Science from Worcester Polytechnic Institute.
Grace Wang's background is centered more heavily on finance, growth and go-to-market work. Public professional records identify her as a co-founder of OpusClip, with education at Peking University and professional experience including ByteDance.
This division of backgrounds is important when discussing invention. The public record does not establish that one founder personally created every component of the platform.
Young Zhao's Contribution
Zhao's most clearly documented contribution is the identification of the creator-content problem and the product direction that emerged from solving it.
His agency experience exposed him to the difficulty of producing video content at scale.
He then helped establish the company in 2022, participated in the development of the initial livestreaming product, and was involved in the decision to convert the clipping feature into a standalone product after observing user demand.
His public interviews also demonstrate that the company adopted a strongly user-driven product-development approach.
Rather than treating the original product concept as fixed, the founders observed actual user behavior and changed the product when the evidence showed that users valued clipping more than livestreaming.
As CEO, Zhao subsequently led the company's product, commercial and strategic development as OpusClip expanded.
Did Young Zhao Invent the AI Technology?
The evidence supports a narrower and more accurate conclusion.
Young Zhao was a founder and a key product-originator behind OpusClip, but the available public record does not establish that he personally invented all of the underlying AI algorithms or wrote the complete production codebase.
OpusClip's AI capabilities have been developed by a broader technical organization.
The company's Opus AI Research division has publicly documented research in video understanding, video editing, video repurposing, multimodal reasoning and related areas.
Researchers associated with Opus AI Research have published work at major AI conferences including AAAI, CVPR and ACL.
One particularly relevant paper is "Video Repurposing from User Generated Content: A Large-scale Dataset and Benchmark," published at AAAI 2025. The work introduced Repurpose-10K, a dataset containing more than 10,000 videos and more than 120,000 annotated clips for studying the long-video-to-short-video repurposing problem.
Another research project, VEU-Bench, was published at CVPR 2025. It studies video-editing understanding across numerous fine-grained editing tasks.
These publications demonstrate that OpusClip's later technical innovation extends well beyond the original product concept and involves a research team.
Jay Wu and the Technical Development of OpusClip
Jay Wu is publicly identified as OpusClip's co-founder and CTO.
His background includes Electrical and Computer Engineering and Computer Science at Worcester Polytechnic Institute.
Wu is also an author on Opus AI Research publications concerning video understanding and video repurposing.
His participation in research such as VEU-Bench and the AAAI 2025 video-repurposing work provides direct evidence of technical involvement in the company's research program.
This makes Wu particularly important when discussing the engineering and research side of OpusClip.
However, even here, the evidence should not be simplified into a claim that Wu personally invented every algorithm used by the platform.
The public research record identifies him as one contributor among a larger technical group.
Grace Wang and the Business Development of OpusClip
Grace Wang is publicly identified as a co-founder and CMO of OpusClip.
Her background includes finance education at Peking University and experience in growth and go-to-market strategy, including work at ByteDance.
Wang's contribution is therefore primarily documented in company growth, marketing, user acquisition, positioning and commercialization rather than as an individual inventor of OpusClip's computer-vision or machine-learning systems.
This distinction matters because founding a technology company involves many different forms of creation.
A founder may contribute the business concept, customer insight, product strategy, commercialization or organizational leadership without personally developing the underlying machine-learning architecture.
Kai-Fu Lee and OpusClip
Kai-Fu Lee is a prominent computer scientist, AI researcher, entrepreneur and venture investor.
He earned a bachelor's degree in computer science from Columbia University and a Ph.D. in computer science from Carnegie Mellon University.
His career includes AI research at Carnegie Mellon, senior positions at Apple, Microsoft and SGI, the leadership of Microsoft Research China, and the presidency of Google China.
In 2009 he founded Innovation Works, which became Sinovation Ventures, where he has served as chairman and CEO.
Lee's documented AI research history includes speech recognition and natural-language technologies. He has also authored numerous academic papers and patents and has spent decades working across AI research, development and investment.
None of those accomplishments, however, establish him as an OpusClip founder.
The public evidence reviewed for this profile identifies Young Zhao, Jay Wu and Grace Wang as OpusClip's founders.
OpusClip's disclosed investor history includes firms such as DCM Ventures, AI Grant, Millennium New Horizons, Samsung NEXT, GTMfund, Alumni Ventures, Fellows Fund, SoftBank Vision Fund 2 and others.
ZhenFund also publicly lists OpusClip in its portfolio.
The available evidence does not establish Kai-Fu Lee or Sinovation Ventures as an OpusClip founder, technical creator, researcher or engineer.
Accordingly, the appropriate conclusion is:
"A direct connection between Kai-Fu Lee and OpusClip could not be independently verified from reliable public sources."
This does not diminish Lee's substantial contributions to artificial intelligence generally. It simply means those contributions should not be attributed to OpusClip without evidence.
How OpusClip's AI Technology Works
OpusClip's current product combines multiple AI-assisted video-production functions.
The platform analyzes long-form video and can identify sections that may be suitable for short-form distribution. It can then create clips, automatically reframe the footage, generate captions, add visual treatments and prepare the resulting videos for social platforms.
The company's current product documentation describes AI clipping, captioning, reframing, B-roll generation, audio enhancement, voice-over and publishing functionality.
OpusClip's ClipAnything system is designed to work with a broad range of video genres, including content with limited dialogue. The company describes the system as using visual, audio and sentiment cues rather than relying solely on transcripts.
The exact production architecture, model weights and complete model orchestration are not publicly disclosed.
Therefore, it would be inaccurate to claim that OpusClip's entire system is based on one particular publicly known large language model or one specific computer-vision architecture.
Highlight Detection
The core problem OpusClip addresses is not simply cutting video.
It is determining which sections of a long video should become independent short-form pieces.
This requires contextual understanding.
A useful clipping system needs to consider the spoken content, the surrounding context, the structure of the conversation, visual information and the beginning and ending boundaries of a potential clip.
OpusClip's public product descriptions state that its AI analyzes video content and identifies highlighting moments before transforming those moments into short-form videos.
Its newer ClipAnything system expands the approach beyond conventional talking-head or transcript-driven content and is designed for videos such as sports, gaming, vlogs, news and other genres.
The company describes ClipAnything as using visual, audio and sentiment signals.
That is different from claiming that the system can objectively determine whether a video will go viral.
Highlight selection is an automated prediction problem; actual social-media performance depends on many factors outside the editing system.
Video Reframing
Automatic reframing is another major component of the platform.
When a landscape video is converted into a vertical short, the system must determine which part of the frame should remain visible.
OpusClip's AI reframe functionality is designed to follow active speakers and moving subjects and adapt video to formats including 9:16, 16:9 and 1:1.
OpusClip has also conducted public research into video editing and reframing.
Its research organization has published work on video-editing understanding, while later research includes work titled "Reframe Anything: LLM Agent for Open World Video Reframing."
The existence of these research projects supports the conclusion that video understanding and reframing are active areas of OpusClip's technical development.
Captions and Transcription
OpusClip automatically generates captions and allows users to edit the resulting text.
The company currently advertises support for numerous languages, including English, German, Spanish, French, Portuguese, Italian, Dutch, Russian, Polish, Indonesian, Ukrainian, Swedish, Turkish, Norwegian, Croatian, Romanian, Slovak, Greek, Danish, Finnish, Hungarian, Czech, Japanese, Korean and Vietnamese, among others.
Captioning should be distinguished from translation and dubbing.
Transcription converts speech into text.
Captioning places that text into timed visual form on the video.
Translation converts the original spoken or written language into another language.
Dubbing or voice-over involves generating or replacing spoken audio.
OpusClip now offers multiple AI-powered audio and language capabilities, but individual features should be evaluated against the company's current documentation rather than assuming that every language feature applies identically across every workflow.
AI Virality Score
OpusClip has marketed an AI Virality Score designed to estimate the potential performance of generated clips.
The existence of a score does not mean that OpusClip can guarantee a video's future views or engagement.
A score should be understood as a model-generated estimate rather than a guaranteed outcome.
The public material reviewed does not establish that the score provides a scientifically validated prediction of future social-media performance across all platforms and audiences.
Therefore, articles about OpusClip should avoid statements such as "the AI predicts which videos will go viral."
A more accurate description is that OpusClip provides an AI-generated estimate intended to help users evaluate generated clips.
Opus AI Research
OpusClip's technical development has expanded into formal AI research.
Its Opus AI Research division describes itself as the company's research division focused particularly on video understanding.
Its publicly listed work includes:
- Video Repurposing from User Generated Content: A Large-scale Dataset and Benchmark
- VEU-Bench: Towards Comprehensive Understanding of Video Editing
- RSVP: Reasoning Segmentation via Visual Prompting and Multi-modal Chain-of-Thought
- Number it: Temporal Grounding Videos like Flipping Manga
- Frame Order Matters: A Temporal Sequence-Aware Model for Few-Shot Action Recognition
- Object-Level Out-of-Distribution Detection via Regional Uncertainty Alignment of Multimodal Representations
- VisTa: Visual-contextual and Text-augmented Zero-shot Object-level OOD Detection
This research shows that OpusClip's technical work is not limited to a consumer-facing editing interface.
The company is also researching datasets, benchmarks, multimodal video understanding and AI systems that can reason about video-editing tasks.
The Wider Technical Team
Public research records identify technical contributors including Yongliang Wu, Wenbo Zhu, Jiawang Cao, Yi Lu, Bozheng Li, Weiheng Chi, Zihan Qiu, Lirian Su, Haolin Zheng and Jay Wu, among others.
Some contributors have documented roles as machine-learning engineers, applied scientists or research interns at OpusClip/Opus AI Research.
For example, Jiawang Cao is publicly identified as an Applied Scientist at OpusClip and has contributed to research involving video repurposing and video reframing.
This evidence reinforces an important point: the AI technology behind OpusClip cannot responsibly be attributed to a single founder.
The product originated with a founding team and evolved through the work of a larger engineering and research organization.
What Did OpusClip Actually Invent?
OpusClip did not invent artificial intelligence, machine learning, speech recognition, computer vision, video editing, automatic transcription, face detection or digital video processing.
These technologies existed decades before the company.
OpusClip's contribution is better understood as product and system innovation around the problem of AI-assisted video repurposing.
The company's platform combines video understanding, content selection, editing, reframing, captioning, formatting and distribution into a creator-oriented workflow.
Its research program also contributes new datasets, benchmarks, models and techniques for video understanding and video editing.
The exact boundary between proprietary product technology and third-party/open research is not completely public.
Therefore, claims that OpusClip "invented AI video editing" or "invented automated video clipping" would be historically inaccurate.
Founder vs Inventor vs Engineer
The history of OpusClip demonstrates why these terms should not be treated as interchangeable.
Young Zhao is a verified co-founder and CEO whose public account documents the problem that inspired the company and the product pivot that created the standalone clipping platform.
Jay Wu is a verified co-founder and CTO with a technical background and documented participation in Opus AI Research publications.
Grace Wang is a verified co-founder and CMO whose background and public role center heavily on growth, marketing and commercialization.
The later technical system was developed by a broader group of AI researchers, applied scientists, machine-learning engineers and software engineers.
Therefore, OpusClip is best understood as a team-created product rather than the invention of one individual.
Current Status in 2026
As of September 2026, OpusClip describes itself as an AI video clipping and editing platform and is increasingly positioning the product as an AI video creation and editing agent.
Its current website reports more than 20 million creators and businesses, more than 440 million clips created and more than 160 billion total views from clips posted by users, with those figures identified as of September 2026.
The current platform supports long-form video clipping, AI captioning, AI reframing, AI B-roll, audio enhancement, AI voice-over, publishing, analytics and workflow integrations.
OpusClip also offers an API intended to allow developers and AI agents to integrate video clipping, captions, reframing and translations into other workflows.
The company's current headquarters is listed in Mountain View, California, with additional offices listed in San Francisco, Vancouver and Los Angeles.
Funding and Growth
OpusClip announced $30 million in cumulative funding in August 2024, including its Series A led by Millennium New Horizons and participation from AI Grant, Samsung NEXT, GTMfund, DCM Ventures, Alumni Ventures, Fellows Fund, Alpine VC and angel investor Jason Lemkin.
In March 2025, the company announced a new investment led by SoftBank Vision Fund 2.
The reported financing and valuation figures should be presented with their respective announcement dates rather than treated as current financial metrics.
The funding history establishes investor participation, but investment should not be confused with invention or engineering.
Who Actually Created OpusClip?
The strongest evidence supports the following conclusion:
OpusClip was founded in 2022 by Young Zhao, Jay Wu and Grace Wang.
Young Zhao was a key product-originator whose experience running a social-media talent agency led him to identify the difficulty of producing video at scale. He and the founding team initially built an AI-powered livestreaming product and later pivoted toward the clipping functionality that users valued.
Jay Wu was a co-founder and CTO and is directly connected to the company's technical research program.
Grace Wang was a co-founder and CMO whose documented contribution centers on growth, go-to-market strategy and commercialization.
The AI systems powering the modern platform were developed by a broader technical team, including researchers and engineers associated with Opus AI Research.
Kai-Fu Lee should not be described as a founder, co-founder or inventor of OpusClip based on the evidence currently available.
The public record therefore supports a team-level account of OpusClip's creation.
There is no reliable evidence establishing a single individual as the sole inventor of the entire OpusClip technology stack.
Evidence Classification
VERIFIED:
- Young Zhao, Jay Wu and Grace Wang are identified as OpusClip founders.
- OpusClip was founded in 2022.
- OpusClip launched its standalone clipping product in June 2023.
- Young Zhao is co-founder and CEO.
- Jay Wu is co-founder and CTO.
- Grace Wang is co-founder and CMO.
- Opus AI Research exists as the company's research division.
- OpusClip researchers have published work at AAAI, CVPR and ACL.
- ZhenFund lists OpusClip among its portfolio companies and identifies the three founders.
COMPANY CLAIM:
- Current user, clip and view counts.
- Product performance claims.
- Claims concerning AI quality and "virality."
- Marketing descriptions of ClipAnything and other proprietary systems.
REPORTED:
- Some funding amounts, valuations and investor details reported by business publications and startup databases.
UNKNOWN / NOT PUBLICLY DOCUMENTED:
- Young Zhao's exact birth date and birthplace.
- The complete internal architecture of OpusClip.
- The exact proprietary models used in every stage of production.
- The identity of one individual who can be credited as the sole inventor of OpusClip.
- A verified founder, engineering, research or advisory role for Kai-Fu Lee at OpusClip.
Final Assessment
The evidence does not support a single-inventor narrative.
The most defensible historical description is that Young Zhao, Jay Wu and Grace Wang founded OpusClip in 2022, with Zhao providing a particularly well-documented account of the product insight and pivot that produced the standalone clipping platform.
The technical capabilities of modern OpusClip are the result of continued work by a broader engineering and AI research team.
Kai-Fu Lee is a major figure in the history of artificial intelligence and venture investing, but a direct connection between Kai-Fu Lee and OpusClip could not be independently verified from reliable public sources. He should therefore not be listed as an OpusClip founder or inventor unless new primary evidence establishes such a relationship.
In short:
Founder: Young Zhao, Jay Wu and Grace Wang
Product origin most clearly documented through: Young Zhao and the founding team
Technical development: Broader OpusClip engineering and Opus AI Research teams
Sole inventor: Not publicly documented
Kai-Fu Lee: No verified OpusClip founder/inventor role established
