Pika Labs
Last reviewed
May 31, 2026
Sources
No citations yet
Review status
Needs citations
Revision
v2 ยท 2,514 words
Improve this article
Add missing citations, update stale details, or suggest a clearer explanation.
Suggest edit
CtrlK
Last reviewed
May 31, 2026
Sources
No citations yet
Review status
Needs citations
Revision
v2 ยท 2,514 words
Add missing citations, update stale details, or suggest a clearer explanation.
Pika Labs (legally incorporated as Mellis, Inc. and doing business as Pika) is an American artificial intelligence company that develops generative video software. Founded in April 2023 by two Stanford University computer-science PhD students, Demi Guo and Chenlin Meng, the company built one of the first consumer-facing text-to-video generators to achieve mainstream attention and has been frequently grouped with Runway and OpenAI's Sora as a pioneer of the AI video category.[1][2][3]
The company is headquartered in Palo Alto, California, and remains privately held. Between April 2023 and June 2024 it raised roughly $135 million across pre-seed, seed, Series A, and Series B rounds, reaching a reported post-money valuation of about $470 million. Lead investors have included Lightspeed Venture Partners and Spark Capital, with angels including Andrej Karpathy, Nat Friedman, Daniel Gross, Adam D'Angelo, and the actor Jared Leto.[4][5][6]
This article covers Pika Labs as a company: its founders, history, funding, and team. For technical details about the company's generative video models and product features, see pika and pika 2 5.
| Item | Value |
|---|---|
| Legal name | Mellis, Inc. (doing business as Pika) |
| Founded | April 2023 |
| Founders | Demi Guo (CEO), Chenlin Meng (CTO) |
| Headquarters | Palo Alto, California, U.S. |
| Industry | Generative AI / video generation |
| Status | Private |
| Total funding | ~$135 million (through Series B, June 2024) |
| Valuation | ~$470 million (June 2024) |
| Lead investors | Lightspeed Venture Partners; Spark Capital |
| Notable product | Pika 1.0 (Nov 2023), Pika 1.5 (2024), Pika 2.0 (Dec 2024), Pika 2.5 (2026) |
Pika Labs was founded in April 2023 by Demi Guo and Chenlin Meng, classmates in Stanford University's AI Lab who both dropped out of their PhD programs to start the company.[1][2][7]
Demi Guo was born in Hangzhou, China, and moved to the United States for higher education. She earned a bachelor's degree in mathematics and a master's degree in computer science from Harvard University, where she was active in entrepreneurship and engineering student groups. She subsequently enrolled in Stanford's computer science PhD program, where she was co-advised by professors Ron Fedkiw and Christopher Manning, working at the intersection of natural language processing and computer graphics.[7][8]
Prior to founding Pika, Guo held research and engineering internships at Google Brain, Microsoft Research, Quora, and Hudson River Trading, and worked as a research engineer at Facebook AI Research (now Meta AI). She was a silver medalist at the 2015 International Olympiad in Informatics and a two-time winner of the USA Computing Olympiad Open division.[7][8] Guo was 25 at the time of Pika's founding and is reported to be one of the youngest founders of a venture-backed generative-AI company to reach a nine-figure valuation.[9]
Chenlin Meng studied mathematics and computer science at Stanford as an undergraduate and continued there for a PhD, advised by Stefano Ermon. Her doctoral research focused on generative models and diffusion-based image and video synthesis. She is a co-author of Denoising Diffusion Implicit Models (DDIM), a widely used sampling method for diffusion models that has been cited thousands of times and is integral to systems including DALL-E 2, Imagen, and Stable Diffusion.[10][11][7]
Before co-founding Pika, Meng interned at Google AI and published on generative AI at venues including NeurIPS, ICLR, ICML, and CVPR. She received the Stanford Interdisciplinary Graduate Fellowship and the Cadence Women in Technology Scholarship.[11][7]
Guo and Meng have said in interviews that the immediate impetus for the company was their own frustration with existing video tools when trying to produce a short film for a Stanford class project: they found the available consumer software too cumbersome and the research-grade systems too inaccessible. They began experimenting with text-to-video models in mid-2022 and 2023 and posted early demos to a Discord server, which quickly attracted hundreds of thousands of users and angel-investor interest.[2][3]
According to multiple interviews, the founders chose the name "Pika" after the small, photogenic mountain mammal of the same name, with the founders citing its visual recognizability and the brevity of the syllables as appealing branding qualities. The company was incorporated in California as Mellis, Inc. but has consistently done business under the Pika and Pika Labs names.[24][4]
Guo and Meng formally took leaves of absence from their Stanford PhD programs to focus on the company; both have stated publicly that they consider their academic enrollment paused rather than abandoned, although neither had returned to the program as of the most recent reporting.[7][9]
Pika Labs raised capital unusually quickly for an AI video startup, completing four announced rounds within its first roughly fourteen months of operation.
In the six months between the company's founding and its public launch in November 2023, Pika Labs raised approximately $20 million across pre-seed and seed rounds. Both rounds were led by the AI-focused angel investors Nat Friedman (former GitHub CEO) and Daniel Gross, investing through their joint fund.[1][4][12]
Other early-round participants included Andrej Karpathy (former Tesla and OpenAI researcher), Elad Gil, Adam D'Angelo (CEO of Quora and a board member of OpenAI), Clem Delangue (CEO of Hugging Face), Aravind Srinivas (CEO of Perplexity), Vipul Ved Prakash (CEO of Together), Mati Staniszewski (CEO of ElevenLabs), Keith Peiris (CEO of Tome), Craig Kallman (CEO of Atlantic Records), and Alex Chung (co-founder of Giphy), as well as the firms Homebrew, Conviction Capital, SV Angel, and Ben's Bites.[1][4][12]
On 28 November 2023, Pika announced a $35 million Series A round led by Lightspeed Venture Partners, bringing total capital raised to approximately $55 million.[1][2][4][12] The Series A was timed with the public release of Pika 1.0, the company's first widely available web product. Lightspeed partners Lisa Han and Guru Chahal led the firm's investment.[13]
On 4 June 2024, Pika announced an $80 million Series B led by Spark Capital, with participation from Lightspeed, Greycroft, and the actor Jared Leto. Bloomberg reported a post-money valuation of approximately $470 million.[5][6][14] Total funding reached approximately $135 million.[6]
At the time of the Series B, multiple sources reported Pika had only around 13 full-time employees, including researchers who had previously worked at Google, Meta, and Uber, an unusually small headcount relative to the size of the round and the company's valuation.[6][14]
As of the publication date of this article, Pika Labs has not publicly announced a Series C funding round. Industry sources have estimated the company's value at higher figures in subsequent secondary-market activity, but no primary-market round has been disclosed.[15]
A full description of Pika's models and product features is maintained at pika. The timeline of major releases attributable to Pika Labs is:
The company also operates standalone tools and mobile apps, including a Pikaffects iOS app and integrations with third-party software such as Adobe Firefly Boards.[20]
Pika Labs has historically operated with a small team. Reporting around the Series B in mid-2024 placed full-time headcount at approximately 13 people, including engineers and researchers with prior experience at Google, Meta, and Uber.[6][14] Third-party data aggregators reported the company had grown to roughly 48 employees by the end of 2024,[21] and over 100 employees by early 2026, though Pika Labs has not publicly confirmed precise figures.
In addition to its institutional lead investors, Lightspeed Venture Partners (Series A) and Spark Capital (Series B), Pika's cap table notably includes a number of prominent AI researchers and operators as angels.
The composition of Pika's cap table, heavy on practitioners in adjacent generative-AI fields rather than purely on traditional financial investors, has been highlighted by industry observers as a distinctive feature of the company's early-stage capital raising, providing both technical guidance and ecosystem distribution channels.[9][4]
Demi Guo serves as chief executive officer and Chenlin Meng as chief technology officer. The pair have not publicly disclosed additional named executive roles such as CFO or COO. Operationally, Guo has spoken about her involvement in fundraising, product, hiring, and external partnerships, while Meng leads research and engineering, including the design and training of successive versions of the Pika video model.[9][7]
Pika Labs occupies a distinctive position in the AI-video market, generally identified as one of the earliest consumer-oriented entrants. Its principal competitors include Runway (Gen-3, Gen-4), OpenAI's Sora, Google DeepMind's Veo line, Kling (Kuaishou), and Luma AI's Dream Machine.[22][23]
Whereas Runway has positioned itself as an editor-centric tool for professional creators and Sora and Veo as quality-leading general-purpose models, Pika has emphasized short-form, social, and prosumer use cases, with a focus on accessible web-based prompting, special effects ("Pikaffects"), and consumer mobile apps.[22][23] Pika 2.5 and earlier versions have been widely reviewed as competitive on short, stylized content but generally behind Sora-class systems on long-duration realism.[19][23]
Industry analysts have described the post-Sora generative-video landscape as bifurcated between very large, generalist frontier-lab models and smaller, more product-focused start-ups; Pika is typically placed in the latter group together with Runway, Luma, and several Chinese entrants such as Kling and Hailuo. Pika's distinguishing competitive strategy as of 2024-2026 has been a heavy emphasis on stylized social-video features (the "Pikaffects" line and templated transformations) and on a hosted web product with a generous free tier and consumer-priced paid plans.[9][22]
Pika has also taken a notably different go-to-market path from Runway, which evolved from a creative-software company already serving professional video editors. Pika launched directly to consumers on Discord and the web, and many of its users came from existing AI-art communities. The company has subsequently added more traditional creator-economy and professional-tooling features (such as the Studio editor in Pika 2.5) and entered into ecosystem partnerships including integration with Adobe Firefly Boards.[20][19]
Pika's November 2023 launch coincided with, and was an early example of, the rapid 2023-2024 emergence of consumer text-to-video AI as a category. Demi Guo and Chenlin Meng have been profiled in mainstream business publications including Bloomberg, Inc., TechCrunch, Forbes, and the South China Morning Post, and have featured on lists of notable women in AI.[6][9][1][7]
The company is occasionally cited as evidence of the lean nature of post-foundation-model startups: a very small team building on top of (or alongside) external research advances can, with concentrated capital, reach significant valuations and large user bases within a year of incorporation.[6][14] Critics have noted that this dynamic also makes such companies vulnerable to capability shifts at the frontier laboratories, as illustrated by industry reactions to OpenAI's announcement of Sora in February 2024.[22]
Pika has also drawn attention because of its founders' demographic profile. Both Guo and Meng are women in their twenties at the time of founding, in a field where founder demographics skew heavily male and older. Multiple "women in AI" profiles and lists, including an early 100 Women in AI feature on Meng and several profiles of Guo in Chinese-language business media, have highlighted the pair as notable examples of young female-led AI start-ups.[11][7][25]
The release of Pika 1.0 in late 2023 attracted considerable mainstream press coverage in part because the company's product demonstrations included stylized "anime," "3D animation," and "cinematic" clips that captured public imagination as AI video moved from research curiosities to user-facing tools. Coverage at the time often paired Pika with Runway and the then-newly announced Stability AI Stable Video Diffusion model.[16][3]
As of the company's most recent product release, Pika 2.5, the company has continued to position itself as a consumer- and creator-focused player rather than a foundation-model laboratory, with public statements from Guo and Meng emphasizing product polish, latency, and accessibility over headline benchmark numbers.[19][9]