HeyGen ARR doubles to $200 million in eight months: 85% of Fortune 100 digital users

While the AI ​​video track was still discussing "model running scores", HeyGen used a set of business figures to complete the dimensionality reduction attack: On June 25, the company announced that ARR had doubled in eight months to reach 200 million. Founder Joshua Xu attributes the growth to a concept - "identity-first AI video": more than 30 million users, customers in 196 countries, and 85% of Fortune 100 companies adopting it.

Growth engine: not "generating videos" but "reproducing identity"

HeyGen's core products are AI digital human/virtual image video (Avatar), AI video translation, lip synchronization (Lip Sync), localization, etc. Different from "generating a video", its value anchor lies in allowing the image and voice of a real person/brand to be reproduced in an AI video - which explains why companies are willing to pay for this: marketing, training, sales, and localization are all scenarios that "require real people to appear", and digital people have completely changed the cost structure of these scenarios.

Intensive period of products and honors

In the first half of 2026, HeyGen’s product updates are intensive:

• April: Release of Avatar V (officially "the world's most realistic AI digital human model")
• April: Introducing Seedance 2.0 (video generation)
• May: G2 Summer 2026 report won 281 badges, 23 first place in categories
• June: HyperFrames skills, Live Avatar API, Speech Cleanup, cloud rendering and other updates

From an industry perspective, $200 million in ARR has pushed HeyGen to the most successful commercialization position in the AI ​​video/digital human track, directly competing with Synthesia and others. But more important than the numbers is its narrative turn: "Identity first" means that the end point of AI video competition is not "picture quality", but "trust" - companies are willing to pay for an image that is "as reliable as a real person". For the domestic digital human track (Tencent Zhiying, Silicon-based Intelligence, Laihua, etc.), HeyGen proves a key conclusion: digital human is not a demonstration product, but a real business that can support hundreds of millions of dollars in ARR. The key is to find the payment needs such as "identity reproduction".

Several directions worth tracking in the future:

1. Sustainability of ARR growth: Is doubling in eight months a sprint or a new normal? Q3/Q4 data is the touchstone.
2. Avatar V’s upper limit of realism: Whether digital humans can cross the "uncanny valley" determines the ceiling for enterprise adoption.
3. Ecological Effect of Live Avatar API: Can the open real-time digital human API generate a third-party application ecosystem?
4. Commercial Benchmarking of Domestic Digital People: Who can take the lead in making "identity reappearance" a scalable payment scenario.