Keling AI

• 1. overview (618 chars)

  Keling AI is a large AI video generation model launched by Kuaishou in June 2024. With its powerful video generation capabilities, it is known as "the world's first real image-level video generation tool open to the public." As of October 2024, the number of users has exceeded 3.6 million, and a total of more than 37 million videos have been generated. In 2025, the Keling AI 3.0 series will be launched, and the native 4K direct-out function will be launched for the first time in the industry, supporting one-click generation of cinema-quality video. In terms of pricing, it adopts the inspiration value membership system. Gold membership is 66 yuan/month. It is currently one of the most competitive AI video generation tools in China.

• 2. Product background (996 chars)

  Keling AI is independently developed by Kuaishou and is one of the earliest open large-scale AI video generation models in China. Kuaishou, as the leading short video platform in China, has been early in the field of video AI. The launch of Keling AI marks Kuaishou's official entry into the AI ​​video generation track.   On June 6, 2024, Keling AI was officially launched and opened for invitational testing, and then opened for public testing and launched a membership system in July. The director co-creation plan will be launched in September 2024, and the number of users will exceed 3.6 million in October 2024. In December 2024, the base model will be upgraded to version 1.6, and an independent App will be launched. It will be fully connected to the DeepSeek-R1 intelligent system in March 2025, launch the Creator Program 3.0 in April 2025, and launch the industry's first native 4K direct video function in May 2025.   Keling AI's technical architecture has been recognized by the industry and was hailed as a "milestone in the global cultural video field" by MIT Technology Review. In authoritative evaluation lists such as VideoGen Arena, Keling AI 3.0 has ranked first in ELO scores.

• 3. Product functions and usage experience (1927 chars)

  Keling AI provides seven core functions, covering all scene needs from basic creation to professional customization.   First, high-definition video generation. Keling AI supports 1080P resolution and 30fps smooth images, and uses a 3D spatio-temporal joint attention mechanism to create movie-level visual effects. The native 4K direct output function launched in May 2025 is the first in the industry to support the ability to generate 4K videos with one click, and can output film-quality images without complex post-processing.   Second, multi-modal creation support. Keling AI supports three modes: Wensheng video, Tusheng video, and video continuation. Users can choose the most suitable creative method according to their specific needs.   Third, first and last frame control. This is the first function of Keling AI. Users can customize the start frame and end frame to achieve silky transitions and one-shot to end special effects. This is a differentiated advantage among similar products.   Fourth, intelligent mirror movement mode. Keling AI provides six master-level lens control solutions, including professional-level film and television languages ​​such as push-zoom and horizontal panning, helping users easily achieve professional camera effects.   Fifth, AI digital human customization. Accurately generate personalized virtual images, support fine adjustment of gender, age, skin color and other characteristics to meet the needs of brand marketing and content creation.   Sixth, video continuation function. It supports a single continuation of 4.5-second videos, and can be continuously superimposed and generated to 3 minutes to meet the needs of long video creation.   Seventh, video face model. This function is currently in the internal testing stage and supports users to independently train personalized face models, effectively solving the problem of character consistency.   Judging from user feedback, Keling AI has performed well in terms of video generation quality, especially in terms of naturalness of character movements and scene consistency, and has received much praise. However, some users have reported that the free version has the problem of too long video generation time, and the membership version has a better experience.

• 4. Pricing and business model (981 chars)

  Keling AI adopts an inspiration value membership system, and the specific pricing is as follows:   Free users receive a certain amount of free inspiration points every day, which is suitable for users with light experience. Gold membership is priced at 66 yuan/month, provides 660 inspiration value, and supports high-definition watermark removal and batch creation privileges. In addition, there are multiple membership services to meet the needs of different users.   From the perspective of business model, Keling AI is mainly monetized through membership subscription and inspiration value consumption. Reports in March 2025 showed that Keling AI’s annual recurring revenue (ARR) reached 340 million yuan and is expected to reach 240 million yuan in 2025, showing good commercialization prospects.   Creator Plan 3.0 is another important ecological strategy. The new plan launched in April 2025 will invest a million-level cash pool and tens of millions of "inspiration value" points every month. Creators will publish original videos on Douyin, Xiaohongshu, Bilibili and other platforms, and can receive up to 6,000 yuan in cash rewards for a single work.

• 5. User reviews (871 chars)

  Judging from public information, users’ evaluation of Keling AI is generally positive.   In terms of positive reviews, users recognized the production quality of Keling AI, especially its amazing performance in the field of AI short dramas. Some users said that "the AI ​​short drama produced by Keling is amazing enough" and believed that it has practical value in the field of short drama. In the 2026 domestic AI video tool evaluation, Keling 3.0 performed well in terms of understanding, stability, and controllability.   Negative feedback mainly focuses on the following points: the waiting time for the free version to be generated is long; there is still room for improvement in character consistency in some scenarios; the pricing is high for high-frequency creators.   In terms of industry recognition, Keling AI has won a number of honors, including the first prize of the expert review of the first "inspiration" burst of Keling AIx Quick Video Creation Competition, "the most immersive AI audio-visual project of the year", etc.

• 6. Industry evaluation (854 chars)

  From an industry perspective, the launch of Keling AI is of landmark significance.   First of all, the emergence of Keling AI has promoted the development of the domestic AI video generation track to a certain extent. As one of the first AI video tools open to the public, its productization capabilities and user scale provide a reference benchmark for the industry.   Secondly, innovation at the technical level is recognized. The native 4K direct-out function is the first of its kind in the industry and is considered to represent an important direction in AI video generation. In evaluations such as VideoGen Arena, Keling AI 3.0 has achieved a leading position.   However, there are also industry views that Keling AI faces fierce market competition. Compared with international competitors such as Runway, Pika Labs, and Luma, as well as domestic platforms such as Byte and Kuaishou, Keling AI needs to continue to innovate in aspects such as production quality, production speed, and pricing to remain competitive.

• 7. Controversies and Risks (626 chars)

  Judging from public information, the controversy about Keling AI is relatively limited and mainly focuses on the following points:   First, content security risks. As a video generation tool, Keling AI may be used to generate false information or infringing content, and the platform needs to continue to strengthen its content review capabilities.   Second, user privacy protection. The AI ​​digital human customization function involves user portrait data, and how to ensure user data security is an important issue.   Third, market competition has intensified. As more players enter the AI ​​video generation track, market competition will become more intense, and Keling AI needs to continue to innovate to maintain its leading position.

• 8. Applicable groups and usage suggestions (818 chars)

  Keling AI is suitable for the following user groups:   First, AI video content creators. Whether it is an individual creator or a professional team, Keling AI's multi-modal functions and intelligent camera mode can provide effective support.   Second, short video and short drama practitioners. Keling AI's outstanding performance in the field of AI short dramas makes it one of the choices for practitioners.   Third, brand marketing users. The AI ​​digital human customization function can meet brand marketing needs.   Fourth, professional users of film, television and advertising. The native 4K direct-out function provides new options for professional video creators.   For users with light experience, it is recommended to use the free version first to understand the product capabilities. For high-frequency creators, gold membership is relatively cost-effective. If you need higher quality 4K output, you can pay attention to professional membership rights.

• 9. Summarize (458 chars)

  Keling AI, as an AI video generation tool launched by Kuaishou, occupies an important position in the domestic market with its productization capabilities and technological innovation. The native 4K direct-out function launched in 2025 will further strengthen its technological leadership. In terms of pricing, the gold membership price of 66 yuan/month has certain competition among similar products. For AI video creators, Keling AI is a choice worth considering. It is recommended to choose based on your own creative needs and budget.

No in-depth report yet

• 

  ztz6fx

  —

  可灵AI的4K直出功能是真的强！之前用其他工具生成完还得自己调分辨率，可灵直接一步到位，省了我太多后期时间。

---

• 

  Barbara.Cooper_Max

  —

  黄金会员66元/月，性价比我觉得还不错。之前免费版生成要等半天，现在会员版基本几分钟就出，质量也更高。

---

• 

  兰花_15

  —

  实测文生视频功能，首尾帧控制真的香！可以自己定义起始和结束帧，做转场特别自然，一镜到底的效果YYDS。

---

• 

  Ashleybarker

  —

  可灵AI 3.0新出的控制功能更强了，能分层控制运镜、场景和人物，之前生成视频总的感觉动作不够自然，现在好很多。

---

• 

  流光

  —

  国内AI视频工具终于站起来了！可灵3.0在VideoGen Arena拿第一不是吹的，生成质量完全不输Runway和Pika。

---

• 

  吕玉平

  —

  我用可灵做短视频，智能运镜模式真的好用，六种镜头方案随便选，推拉摇移跟专业摄影师拍的似的。

---

• 

  Joseph.Jackson08

  —

  刚升级了3.0版本，不得不说人物一致性问题改善太多了！之前生成连续帧总会出现人脸崩坏，现在基本不会。

---

• 

  Jerry.Phillips_202420

  —

  可灵的免费版每天给的那点灵感值完全不够用，想做长视频的还是要买会员。

---

• 

  胡月云

  —

  创作者计划3.0太香了！一个月投入百万现金池，单条视频最高奖6000块，我已经领了好几笔了。

---

• 

  Nathan.Sanders520

  —

  可灵AI的数字人定制功能好使！可以自己调节性别年龄肤色，做品牌视频很方便。

---

• 

  Ralph.Alvarez

  —

  生成速度比之前快了不少，但跟Runway比还是有点差距，高频创作者可能不够用。

---

• 

  Ruth_Torres520

  —

  用可灵做了十几个AI短剧，播放量最高的破百万了！虽然还有一些瑕疵，但这个价格还要啥自行车。

---

• 

  JennaJenkins

  —

  视频续写功能最多可以生成3分钟，这个长度做短视频足够用了。

---

• 

  Heather_Perry_X89

  —

  可灵AI的AI人脸模型还在内测，期待能早点用上，这样生成连续人物视频就更方便了。

---

• 

  EmilyJackson_9979

  —

  说实话，免费版体验确实一般，等太久。但会员版性价比还是OK的，66块钱一个月还要啥自行车。

---

• 

  Bobby_Mitchell0078

  —

  3D时空联合注意力机制确实有点东西，画面质感比1.0版本好太多。

---

• 

  MsOleaSommervold_88

  —

  接入了DeepSeek-R1之后，智能程度提升了不是一点半点，理解提示词更准了。

---

• 

  CharlesPhillips

  —

  可灵AI的1080P/30fps参数不是盖的，生成出来的视频直接就能用，了很多后期。

---

• 

  郭雅

  —

  在AI视频领域，可灵算是国内头部的存在了。希望以后能更便宜点。

No reviews yet