Video AIGC "King" Pika Emerges, Valued at $200 Million in Six Months
TechFlow Selected TechFlow Selected
Video AIGC "King" Pika Emerges, Valued at $200 Million in Six Months
Behind it is a Chinese team of only four people.
Navigating Web3 tides with focused insightsBy Mu Mu
Backed by a Chinese founding team, valued at over $100 million, and with participation from OpenAI co-founders, the AI video generation tool Pika 1.0 burst onto the scene with full momentum. Within days, it went viral across the internet, amassing over 550,000 community users. Netizens have hailed it as the best video generation tool currently available—emerging as the strongest competitor to Runway’s Gen-2.
Pika Labs, the company behind Pika 1.0, was founded in April this year with the mission of “empowering everyone to become a creative video director and producer.” Currently, Pika 1.0 can generate various types of videos including 3D animations, anime, and cinematic content, while also supporting advanced editing features such as canvas extension, localized edits, and video duration expansion. Real-world testing shows that Pika 1.0 outperforms Gen-2 in generating cinematic shots.
In just a few months, Pika Labs has secured $55 million in funding, achieving a valuation exceeding $200 million. The investor roster reads like a who's who of the AI world, including OpenAI co-founder Andrej Karpathy and Hugging Face co-founder & CEO Clem Delangue.
More notably, Pika Labs is a lean team of only four members—all of Chinese descent. Founder Wenjing Guo and co-founder & CTO Chenlin Meng were both Ph.D. students at Stanford’s AI Lab; third founder Karlin Chen holds a Master’s degree in ML&CV from CMU and previously worked as an engineer at SenseTime; and the fourth team member, Matan Cohen Grumi, serves as Creative Director and is also a television commercial director.
Interestingly, the reason founder Wenjing Guo and her team built Pika 1.0? Because Gen-2 wasn’t good enough.
Pika 1.0 Excels at Cinematic Videos
Beyond standard AIGC capabilities like text-to-video and image-to-video generation, Pika 1.0 enables video style transfer, direct content editing, and aspect ratio changes. These features shift AI’s role in video from mere “generation” to true “production.”
According to official demo videos, users input text describing characters, scenes, and styles, and the tool generates corresponding video footage. For example, entering “Elon Musk in a spacesuit, 3D animation” produces a vivid animated version of astronaut Musk, complete with a successfully launched rocket in the background—the character design, movements, and environment all rendered with striking realism.
In addition to text-to-video, Pika 1.0 supports image-to-video generation: upload an image along with prompt text, and the static image comes alive according to your description.
Want to change a video’s artistic style? Simply enter a style prompt—whether anime or pixel art—and seamlessly blend different visual aesthetics into a single “Everything Everywhere All At Once”-style video.
Pika 1.0 also allows for precise video editing: select a region and provide a text prompt to modify only that area. For instance, draw a box around a model’s clothing to instantly change its design and color—or put sunglasses on a gorilla with one click.
Additionally, Pika 1.0 enables effortless video resizing—a feature already seen in AI image tools like Midjourney, but now pioneered first in video by Pika.
Netizens compared Gen-2 and Pika 1.0 using the same wedding scene prompt, concluding that Pika 1.0 better leverages cinematic techniques to enrich video depth. While Gen-2 delivers higher visual quality, it exhibits clear flaws in rendering human walking motions.
After watching Pika’s promotional video, Hugging Face CEO Clem Delangue remarked: “In 2024, mainstream media will be flooded with AI-generated content.”
Currently, Pika 1.0 is free to use, though access requires an invitation. One netizen vividly captured its current popularity:🔽
In short, Pika 1.0—offering free trials—is now in a red-hot state where access is highly sought after.
550K Users Generating Millions of Videos Weekly
Although Pika 1.0 excels in cinematic output, it wasn't designed specifically for professional filmmaking. “We’re not building a film production tool—we’re creating a product for everyday consumers,” explains Pika CEO Wenjing Guo. “Even though we’re creative, we’re not professionals. But if we had a tool like Pika, we’d likely shine at an AI film festival.”
Guo Wenjing’s journey with AI film festivals dates back last winter, when Runway—an AI startup then valued at $500 million—hosted its inaugural AI Film Festival with a $10,000 prize. Guo teamed up with fellow Stanford classmates to spend their winter break creating a film using generative AI, confident they could win.
However, lacking formal film training and despite using Runway’s AIGC tool Gen-2, their final video fell short. They missed the award, leaving Guo deeply disappointed.
That disappointment became the catalyst for Pika: since existing tools weren’t up to par, they decided to build their own. In April, Guo and her classmate Chenlin Meng dropped out of Stanford to found Pika.
Wenjing Guo (left) and Chenlin Meng
These two powerhouse women wasted no time diving into their venture. Three months later, Pika launched a Discord server similar to Midjourney’s interface—users generate short videos by typing prompts or uploading images in chat channels, then share them within the community.
Quickly, tens of thousands of users flocked to the Discord server. Today, it hosts 550,000 members and continues growing rapidly, producing hundreds of thousands of new videos every week.
This four-person startup has also swiftly completed three funding rounds totaling $55 million. The first two were led by GitHub’s former CEO Nat Friedman, while the latest $35 million Series A round came from Lightspeed Venture Partners. According to Forbes, Pika Labs is now valued between $200 million and $300 million.
Pika’s investor list includes some of Silicon Valley’s top AI figures: OpenAI co-founder Andrej Karpathy, board member Adam D'Angelo, Hugging Face co-founder & CEO Clem Delangue, and Perplexity CEO Aravind Srinivas.
Michael Mignano, partner at Lightspeed, invested in Pika in September and praised the small team: “For startups, speed is the biggest advantage—and this team is the fastest I’ve ever seen.”
Nat Friedman, former GitHub CEO and one of Pika’s investors, also marveled at their efficiency. Reportedly, he suggested adding a text-embedding-into-video feature one afternoon—and by 3 a.m. that night, he was told the feature was already live.
Currently, Pika hasn’t disclosed many technical details about its underlying models, but its speed and product quality have already excited investors. Yet, under the spotlight, Pika still faces intensifying competition.
Just last week, Runway unveiled a new motion slider (“dynamic brush”) feature—swipe over an image to animate it. Beyond consumer offerings, Runway has also begun collaborating with film studios.
Meanwhile, the most popular text-to-image tool, Midjourney, is actively developing video generation capabilities.
Stability AI has already released its Stable Video Diffusion video model, allowing users to adjust parameters like inference steps and denoising strength to control aspects such as style, pose, and line work during generation. SD also supports 3D object synthesis.
The AIGC video space is blossoming with innovation—now it all comes down to who has the most creativity.
Join TechFlow official community to stay tuned
Telegram:https://t.me/TechFlowDaily
X (Twitter):https://x.com/TechFlowPost
X (Twitter) EN:https://x.com/BlockFlow_News
