TL;DR
- Board review: Kuaishou confirmed a review of KlingAI’s board of directors that could bring in outside capital into the video division.
- Evaluation conditions: The $2 billion round being discussed would value Kling at nearly $20 billion, but Kuaishou did not confirm the terms.
- Revenue scale: Kling’s annual sales are discussed at $500 million, roughly double the level before the Lunar New Year.
- Product interests: Due to Kling’s computing needs, creator workflows, and ByteDance’s competition, the product’s speed is at the center of our review.
Documents filed by Chinese internet company Kuaishou on Tuesday document a restructuring of its Kling AI video business, with the board evaluating a possible overhaul that could bring in outside capital to the unit. Investors were quick to react, with Kuaishou shares rising as much as 10% on Tuesday morning after Kling’s new plans were circulated.
Kling AI is Kuaishou’s generative AI video and creative tools business. This includes AI models and products for creating or editing videos, images, sounds, and multimodal content from prompts and reference materials.
Kling is no longer just a subject of outside speculation. Kuaishou did not disclose the final deal, investor list or form of the spin-out, but acknowledged that a different structure may be needed as the business grows. For now, the filing will give investors an official review to follow, rather than new market rumors.
This review is tied to a product category built around a creator’s workflow, rather than a simple social feed feature. Users expect quick control, editable output, and reproducible results that can be adapted to marketing, entertainment, or social video projects. The role of Kling products helps explain why investors are treating Kling as a business line with a unique growth profile.
Kuaishou simply records the review itself. The board is reviewing Kling, which may involve external capital from the division. Valuation discussions and the names of investors remain private to the company.
Why the story changes depending on the application
A board review moves the discussion from rumors to governance. Directors aren’t just deciding whether Kling needs more money. They are deciding whether the video business should continue to grow as part of the parent company or take a more direct route to outside support.
One scenario currently circulating shows Tencent participating in a round of approximately $2 billion at a valuation of $20 billion. This level of valuation is significantly higher than Runway’s reported valuation and is close to Kuaishou’s own market capitalization. Although Kuaishou does not back up these numbers with its own disclosures, the numbers help explain why the filing quickly attracted investor attention.
Kuaishou currently has three major options to consider. The choice is to keep Kling completely in-house, bring in outside help while maintaining control, or move the business to more independent means. Each route will affect management authority, transparency of transactions, and how much outsiders can learn about Kling’s own economics.
Valuation and earnings signals
The stock reaction on Tuesday wasn’t just about headline moves. Investors were reacting not just to Kuaishou’s broad statements that he was interested in AI, but to the possibility of a more direct way to value Kling. Public shareholders can price options on fast-growing AI products before binding conditions arrive, allowing Kuaishou to undergo immediate market testing before term sheets are disclosed.
Discussion of Kling’s own earnings adds another data point. Current discussions about the sector include a revenue run rate of US$500 million per year, about double the pre-Lunar New Year level. Annual occupancy rates are not audited for full year earnings. Predict your current pace for one year. Even with such caveats, the numbers take Kling far beyond a small-scale internal experiment and help explain why shareholders are starting to separate Kuishou’s established short video business from Kling’s growth profile. Products with this kind of revenue discussion can result in clear valuation discussions within much larger consumer internet companies, especially if management later provides investors with clearer numbers about revenue, costs, and user demand.
High valuation scenarios and visible revenue pace make execution more visible. A young AI unit could command a hefty price tag if investors believe growth will continue, but the gap between run-rate revenue and multibillion-dollar valuation puts pressure on Kling to convert user interest into regular creator activity, paid workflows, and enough margin to justify the scale investors are talking about. If an investor believes the asset may merit a different valuation lens, a board-level review could potentially move the stock before final terms exist. It also gives analysts a reason to weigh Kling’s potential separately from Kuaishou’s mature short video advertising and live streaming businesses.
Fast-growing AI products may come with a different risk profile than the parent company’s established platform. Future disclosures will likely provide investors with a benchmark to measure whether Kling is growing to its rumored price, with revenue pace, user adoption, and product expansion alongside key numbers. Analysts will also be looking at whether Kling’s revenue primarily comes from consumer usage, creator subscriptions, enterprise media operations, or platform partnerships. These combinations mean margins and growth durability are different, even if top-line occupancy rates appear to be the same. User composition will also influence how Kuaishou describes the division to shareholders. This is because services for creators are evaluated differently than functions for consumers, which are supported by advertisements. A clearer breakdown will help investors distinguish between long-term use and short-term launch enthusiasm.
How Kling AI economics increase risk
Kling’s resource lawsuit focuses on computing power and talent as AI video competition intensifies. Training upgrades require larger runs before creators can see better clips, but frequent use increases inference costs each time a user generates a video. Product growth may also require workflow tools, localization, moderation, support teams, and customer-facing infrastructure.
AI video generation is especially demanding because each clip combines model inference, frame consistency, motion control, and rendering efforts. While text chatbots can provide short answers in seconds, video tools must maintain visual consistency across many frames. Increased usage may result in a different cost curve than regular social video functionality. For users, these costs appear as everyday product limitations. Slow cues, failed rendering, inconsistent characters, weak editing controls, etc. can be just as important as the raw model behind the clip. With external help, Kling will have more room to add compute, push model updates faster, and support the depth of workflows required for repeated creator use. Professional users don’t just want one impressive demo clip, they also want predictable output over multiple attempts.
Competitive pressures and early strategic signals
Kuaishou is making that choice while competing with ByteDance for short videos. ByteDance’s March 26 expansion of Dreamina Seedance 2.0 audio and video generation with CapCut demonstrated how China’s leading platform groups are rapidly turning video AI for creators into a product competition rather than a feature experiment.
ByteDance is also expanding its Seed AI team into multiple regions. While CapCut provides ByteDance with a distribution channel for creator tools that already exist within editorial workflows, Seed AI’s staffing refers to the model work behind these products. Taken together, these moves demonstrate why rival video platforms can’t treat generated video as a small add-on if they want creators to build recurring habits around their tools. Distribution, model development, and editing workflows all reinforce each other within this category, as users can continue to discover, test, and edit features within the same product environment.
In November 2025, founder and CEO Chen Yishao connected Kling to AI-powered film and television production. Kling is closer to demanding creator and media workflows where scene composition, editing habits, and repetition are as important as novelty. Creators who have built their template, asset, or project habits around one tool are less likely to switch quickly if the service fits into their production routine. For Kuaishou, this kind of workflow lock-in can support revenue beyond one-off experiments, as creators may continue to pay for tools that suit their production habits. Movie and TV use cases put Kling in a more promising category than casual social clips. Users may require multiple versions of scenes, consistent styles across assets, and export paths that fit into existing editing work.
