Pika Labs

Pika Labs (legal name Mellis, Inc., doing business as Pika) is an American generative AI company that builds consumer and creator software for making short AI videos, images and audio. It was founded in April 2023 by Stanford computer-science PhD students Demi Guo (CEO) and Chenlin Meng (CTO), raised about $135 million by June 2024, and was valued at $470 million after its Series B that month.^[1][2][3][4] Fortune reported in October 2025 that Pika had 16.4 million users across its apps.^[5]

Pika was an early consumer-facing text-to-video generator: it had more than 500,000 users of its Discord beta when it announced Pika 1.0 and a new web app in November 2023, into a market that also included Runway and Stability AI.^[2][1][6] The company is based in Palo Alto, California, and is privately held.^[2][7] Between April 2023 and June 2024 it raised pre-seed, seed, Series A and Series B rounds led by Nat Friedman and Daniel Gross, Lightspeed Venture Partners and Spark Capital, with angels including Andrej Karpathy, Adam D'Angelo, Elad Gil and the actor Jared Leto.^{[8][2][9][10]} No later funding round had been announced as of October 2026, and Semafor still described the company as valued at $470 million in September 2026.^[11]

In 2026 the company widened its scope beyond its own video models. It launched "AI Selves," persistent personal AI agents, in February 2026; a developer API marketplace and its first in-house audio models in August 2026; a rebuilt creative platform that routes work across its own and third-party models (including Seedance and MiniMax H3) in September 2026; and a plugin for ChatGPT in October 2026.^{[12][13][14][15][16]}

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.

Key facts

Who founded Pika?

Pika Labs was founded in April 2023 by Demi Guo and Chenlin Meng, PhD students in Stanford University's Artificial Intelligence Lab who left their programs to start the company.^[1][5] The company's June 2024 funding post said that "it's been a year since we dropped out of Stanford to build Pika."^[3] Lightspeed's Series A announcement also named Karli Chen as the third co-founder and founding engineer, a Carnegie Mellon machine learning and computer vision graduate who had worked as a machine learning engineer at SenseTime.^[10]

Demi Guo (co-founder and CEO)

Demi Guo was born Guo Wenjing in Hangzhou, China, according to VnExpress.^[17] She earned a bachelor's degree in mathematics and a master's degree in computer science from Harvard University, then enrolled in Stanford's computer science PhD program, where Lightspeed said she was co-advised by Ron Fedkiw and Christopher Manning; her research sat at the intersection of natural language processing and computer graphics.^[17][10]

Before Stanford, Guo worked at Meta's AI research division (then Facebook AI Research); Pika's 2023 press release said she was its youngest full-time employee, hired as a college sophomore.^[1][2] While at Harvard she was a co-author of the Facebook AI Research protein language model paper "Biological structure and function emerge from scaling unsupervised learning to 250 million protein sequences" (PNAS, 2021), and she was first author of "Parameter-Efficient Transfer Learning with Diff Pruning" with Alexander Rush and Yoon Kim.^[18][19] Lightspeed listed internships at Quora, Google Brain, Microsoft, Tencent and ByteDance.^[10] She won a silver medal at the 2015 International Olympiad in Informatics, and Lightspeed also credited her with first-place finishes in the USA Computing Olympiad.^[17][10] Fortune described her as 27 years old in October 2025.^[5]

Chenlin Meng (co-founder and CTO)

Chenlin Meng studied mathematics at Stanford as an undergraduate (B.S. with distinction, 2016 to 2020) and began a computer science PhD there in September 2020, advised by Stefano Ermon.^[20] Her doctoral research focused on generative models and diffusion models. She is a co-author, with Jiaming Song and Ermon, of Denoising Diffusion Implicit Models (DDIM), a faster sampling method for diffusion models that Pika's 2023 press release said had been widely used in OpenAI's DALL-E 2, Google's Imagen and Stability AI's Stable Diffusion.^[21][2] She is also first author of SDEdit, an image synthesis and editing method, and of "On Distillation of Guided Diffusion Models," written with co-authors including Robin Rombach, Diederik Kingma, Jonathan Ho and Tim Salimans and listed as a CVPR 2023 award candidate; her homepage says the distillation approach was used in Imagen Video.^[22][23][20]

Meng published at venues including NeurIPS, ICLR, ICML and CVPR, and received the Stanford Interdisciplinary Graduate Fellowship and the Cadence Women in Technology Scholarship in 2022.^[20] Lightspeed said she had worked at Google Brain and Stability AI before Pika.^[10] A 100 Women in AI profile described her PhD as "on leave."^[24]

What is the origin story of Pika?

According to Lightspeed, both founders had creative ambitions before turning to computer science: Guo wrote poetry, and Meng wanted to become an animator.^[10] VnExpress reported that the idea for Pika came from Guo's frustration after she and several Stanford PhD classmates entered an AI film festival in New York and did not win, which led Guo and Meng to try to build an easier AI video tool.^[17] Fortune similarly reported that the pair experimented with making long-form generative AI films before deciding to build an easier-to-use video generator.^[5]

The first version of Pika launched as a beta on Discord in late April 2023.^[2][25] By November 2023 the company said it had more than 500,000 users generating millions of videos each week, and that the #pikalabs hashtag had nearly 30 million views on TikTok.^[2]

The company operates under the Pika name; its terms of service identify the operating entity as Mellis, Inc., with a mailing address on High Street in Palo Alto.^[7]

How much funding has Pika raised?

Pika completed four announced rounds within roughly fourteen months of its founding, reaching about $135 million in total funding.^[9][3]

Pre-seed and seed (2023)

Before its public launch in November 2023, Pika raised about $20 million across pre-seed and seed rounds led by Nat Friedman (former GitHub CEO) and Daniel Gross.^[8][2]

The company's November 2023 announcement listed other early investors including Elad Gil, Adam D'Angelo (founder and CEO of Quora), Andrej Karpathy, Clem Delangue (CEO of Hugging Face), Craig Kallman (CEO of Atlantic Records), Alex Chung (co-founder of Giphy), Aravind Srinivas (CEO of Perplexity), Mateusz Staniszewski (CEO of ElevenLabs) and Keith Peiris (CEO of Tome), as well as the firms Homebrew, Conviction Capital, SV Angel and Ben's Bites.^[6][2]

Series A: $35 million (November 2023)

On 28 November 2023, Pika announced a $35 million Series A led by Lightspeed Venture Partners, bringing total capital raised to $55 million.^[1][2] The round was timed with the release of Pika 1.0 and a new web app.^[2] Lightspeed partner Michael Mignano was quoted in the announcement, and VnExpress, citing a Bloomberg interview, described him as a Pika board member; Lightspeed's portfolio page lists partners Lisa Han and Guru Chahal on its Pika team.^[2][17][26]

Series B: $80 million (June 2024)

On 5 June 2024, Pika announced an $80 million Series B led by Spark Capital, bringing total funding to $135 million. The company named its new investors as Greycroft, Lightspeed, Neo, Makers Fund, Jared Leto and Craig Kallman, and listed earlier investors including Mike Vernal, Karim Atiyeh and Aaron Levie alongside those from previous rounds.^[3] The Decoder reported that Pika's valuation rose to $470 million after the round.^[4]

At the time of the Series B, Pika said its team had grown "from three to thirteen," and Maginative reported that the 13 people included AI researchers from Google, Meta and Uber.^[9][3] Maginative also reported that Leto's band, Thirty Seconds to Mars, had used Pika-generated video in its concert footage.^[9]

Later activity

As of October 2026 Pika had not announced a funding round after its Series B. A September 2026 Semafor report still described the company as valued at $470 million.^[11] Some data aggregators list other figures (Sacra, for example, cites reports of a possible valuation of up to $700 million), but these are not tied to an announced round.^[27]

In August 2025, a Stock News roundup published by Tiger Brokers said Meta was reportedly in discussions with Pika about a collaboration that could include an acquisition or a license to its technology; no deal was announced.^[28]

What products does Pika make?

A full description of Pika's models and product features is maintained at the Pika article. The timeline of major releases attributable to Pika Labs is:

When Pika announced version 2.2 on 27 February 2025, the company described it on its official account: "Pika 2.2 is HERE, with 10s generations, 1080p resolution, and Pikaframes, key frame transitions anywhere from 1-10s."^[31] In June 2025 Adobe added Pika 2.2 text-to-video to Firefly Boards alongside Google's Veo 3 and Luma AI's Ray2.^[32] The company's iPhone app, published under the Mellis, Inc. developer name, offers tools including Pikaswaps, Pikamemes, Pikadditions and Pikaffects.^[37]

What are Pikaffects?

"Pikaffects" are pre-set visual effects, introduced with Pika 1.5 on 1 October 2024, that apply physics-defying transformations to an object in a photo or clip. The six launch effects were Melt it, Inflate it, Explode it, Squish it, Crush it and Cake-ify it; the last turns an object into a cake that a knife then slices.^[29] The current iPhone app lists Pikaffects such as Cakeify, Melt, Explode, Squish and Levitate.^[37]

AI Selves and real-time agents (2026)

On 20 February 2026 Pika introduced "AI Selves," which it described as "AI you birth, raise, and set loose to be a living extension of you," with persistent memory, at the separate site pika.me.^[12] In a TBPN interview that month, Guo described an AI Self as "a living extension of you" that is "24/7 available," and said people with expertise could let others talk to their AI Self and monetize that access.^[34]

On 2 April 2026 the company released PikaStream1.0, which it described as a real-time engine that lets an agent hold a face-to-face video call; users could invite their AI Self to a Google Meet, and other agents could use a skill published on GitHub. Pika said the system generates personalized video at 24 frames per second and 480p on a single H100 GPU, with about 1.5 seconds of speech-to-video latency, using a 9-billion-parameter diffusion transformer distilled from a bidirectional teacher into a causal autoregressive student, plus a Transformer-based streaming VAE it calls FlashVAE.^[35] On 10 April 2026 Pika said AI Self owners could earn tokens redeemable for cash when other people talk to their agents or use skills they built; Guo said, "The entire history of AI has been about what AI costs you. We wanted to build the first agent that makes you money."^[38]

In late May 2026 the company opened "Pika Experiments," a public set of open-source agent prototypes beginning with a generative UI demo, and released four marketing skills for its Pika MCP (Model Context Protocol) integration aimed at technical founders.^[39][40]

Pika API Club and audio models (August 2026)

On 5 August 2026 Pika launched the Pika API Club, a $10-per-month membership that gives developers one API for more than 100 video, image, audio, speech and language models, including Seedance 2.0, MiniMax H3, Kling 3.0, Veo 3.1, Grok Imagine Video, GPT Image 2, ElevenLabs, Claude and Gemini as well as Pika 2.5. The company claimed member prices were up to 88 percent lower than other API aggregators.^[13]

On 14 August 2026 Pika announced Pika Audio, which it called its first family of foundation sound models: Pika Soundtrack (video-to-audio), Pika Music, Pika SFX and Pika Speech, initially available only through the API Club and, the company claimed, up to 20 times cheaper than comparable audio models.^[14] According to Pika, Pika Speech is a 3-billion-parameter flow-matching transformer that produces 48 kHz speech, clones a voice from a few seconds of reference audio, handles requests up to five minutes long, and was trained on 403,000 hours of filtered speech, including English and Chinese corpora; the company reported a real-time factor of 0.02 for three-minute requests in its own tests.^[36] Pika said Pika Music can generate tracks up to six minutes long, and Pika SFX up to 20 seconds of 44.1 kHz stereo audio.^[14]

The new Pika and ChatGPT plugin (September and October 2026)

On 17 September 2026 Pika relaunched pika.art as what it called "an AI creative platform made for and by Creatives," combining video, image and sound tools in one place with more than 25 creative apps. The apps automatically choose among the company's own models and third-party ones such as Seedance, GPT Image and MiniMax H3, though users can pick a model themselves; the Video Studio app makes clips or multi-shot videos up to 30 seconds long with sound. Pika said Seedance generations on the platform cost half as much as on other creation platforms, and kept the previous interface at old.pika.art during the transition.^[15] Semafor reported that the product automates which model creates each element of a video and turns basic user input into an editable shot list, and that the launch included apps for relighting and for swapping characters or outfits.^[11]

On 16 October 2026 Pika released a plugin that brings its 40+ creative apps and 100+ models into ChatGPT on desktop and web, with inline tools, a side-panel workspace and a timeline editor; Pika said it worked with OpenAI on the integration.^[16]

On 29 October 2026 Pika released Pika H3 Flex, an open-weight video and audio model that it built by post-training MiniMax H3 to generate in six sampling steps instead of 50.^[42] The open release consists of two LoRA adapters and inference code, distributed under the MiniMax H3 Community License Agreement as "Model Derivatives of MiniMax H3", and covers native 480p and 768p output with synchronized stereo audio. Upscaling to 1080p, 2K and 4K and separation of the soundtrack into five audio stems are hosted features of the Pika API.^[42][43]

Team and notable backers

Pika has operated with a small team. Around the Series B in mid-2024 the company had 13 people, including researchers with prior experience at Google, Meta and Uber.^[9][3] Pika has not published later headcount; the data aggregator GetLatka estimated about 105 employees as of February 2026.^[41] Semafor identified Lindsay Brillson as Pika's head of brand and content in September 2026.^[11]

In addition to its institutional lead investors, Lightspeed Venture Partners (Series A) and Spark Capital (Series B), Pika's cap table includes a number of prominent AI researchers and operators as angels.^[2][3]

• Researchers and AI operators: Andrej Karpathy; Aravind Srinivas (CEO of Perplexity); Clem Delangue (CEO of Hugging Face); Mateusz Staniszewski (CEO of ElevenLabs); Keith Peiris (CEO of Tome).^[6][2]
• AI-focused investors: Nat Friedman (former GitHub CEO); Daniel Gross; Elad Gil; Adam D'Angelo (CEO of Quora).^[2][3]
• Other backers: the actor Jared Leto, Atlantic Records CEO Craig Kallman, Giphy co-founder Alex Chung, Mike Vernal, Karim Atiyeh and Aaron Levie; firms Greycroft, Neo, Makers Fund, Homebrew, Conviction, SV Angel and Ben's Bites.^[2][3]

Leadership

Demi Guo serves as chief executive officer and Chenlin Meng was named chief technology officer in the company's 2023 announcement.^[2] VnExpress reported that Guo oversees product strategy, fundraising and ecosystem development.^[17]

How does Pika compare to Runway and Sora?

Pika has generally been described as one of the earliest consumer-oriented entrants in AI video. Its 2023 launch release said that while other AI video tools focused mainly on professionals and commercial use, Pika was designed for "the everyday consumer and creator."^[2] When it raised its Series B, Maginative listed its competitors as OpenAI's Sora, Google's Veo, Runway, Synthesia and Kuaishou's Kling.^[9]

In October 2025 Fortune contrasted Pika with OpenAI's Sora, which it described as focused on cinematic, photorealistic generation, and Meta's Vibes feed, saying Pika positioned itself closer to the short-form habits of Generation Z and Generation Alpha users.^[5] By September 2026 Semafor wrote that Pika had been "playing catch up" with other AI content-creation companies such as Higgsfield AI, which it reported was valued at $5.4 billion, and noted that Higgsfield and Runway already had integrated creator platforms. Pika pitched its relaunched platform partly on price, saying some features cost half as much as on competing platforms.^[11]

Pika launched directly to consumers on Discord and the web, and later distributed its model through partners such as Adobe Firefly Boards.^[2][32] In 2026 it moved toward an aggregator model, offering third-party video, image and audio models next to its own through its creative platform and API.^[13][15]

Reception and impact

Pika's November 2023 launch came about a week after Stability AI announced Stable Video Diffusion, and coverage at the time described the startup as a direct competitor to Runway and Stability AI.^[1][6] Its launch release described it as a "female-founded AI company," and Meng was later profiled by 100 Women in AI.^[2][24] VnExpress reported that the Pika 1.0 launch also lifted shares of Sunyard Technology, a Hangzhou IT company formerly chaired by Guo's father, by more than 20 percent in one week.^[17]

Its $80 million Series B came when the company had 13 people.^[9][3] Around that round, Guo told The Washington Post, as quoted by The Decoder, "We're very confident that we're building the best foundational model for video."^[4]

What is Pika's strategy for 2025 and 2026?

Through 2025 Pika concentrated on social, consumer-first AI video. Fortune reported in October 2025 that Pika had reached 16.4 million users across its creative apps and had launched a TikTok-like AI video app, followed by a feature called Predictive Video that lets a user upload a selfie and give a simple prompt such as "make me a rock star" or "I'm giving a TED Talk," after which the tool infers intent and generates a complete video with a script, music, dance moves, background, lighting, camera angles and visual effects.^[5] In the same profile, Guo framed the company's bet on casual creators: "Most nonprofessionals will never try to create a film using generative AI, but lots of people like to make short videos," adding that "it's really about self-expression."^[5]

In 2026 the company first focused on AI Selves and agent features, including real-time video and agent monetization.^[12][35][38] From August it turned toward professional creators and developers, with the API Club, its own audio models, the rebuilt pika.art platform and the ChatGPT plugin.^{[13][14][15][16]} Its stated pitch for the September 2026 platform was that "outstanding AI output should be available to people with the taste and ideas to use it," without requiring users to study every model or subscribe to every platform.^[15]

Continue reading

• Pika (video generation)Pika is an artificial intelligence video generation platform developed by Pika Labs, Inc. that lets users create and edit short videos from text prompts, images, and existing clips, and it is best known for…Mentioned 117 times above
• DDemi GuoDemi Guo (born around 1998) is a Chinese-born American technology entrepreneur and former artificial intelligence researcher who is the co-founder and chief executive officer of PikaMentioned 6 times above
• Runway (company)Runway is a New York based artificial intelligence company that builds generative AI tools for video, with its Gen series of models and its GWM-1 world models.Mentioned 5 times above
• Pika 2.5Pika 2.5 is a generative video model developed by Pika Labs, the San Francisco based AI video startup co-founded in April 2023 by Stanford AI Lab dropouts Demi Guo and Chenlin Meng.Mentioned 4 times above
• AAI Video GenerationAI video generation is the use of artificial intelligence systems, predominantly diffusion transformers, to create video clips from text descriptions, still images, or other video, producing sequences of…Listed in See also
• Higgsfield AIHiggsfield AI is an American artificial intelligence company that develops video and image generation software for social media creators, marketers, and enterprise content teams.Listed in See also

More from See also

• List of neolabs
• Neolab

More to read

• Generative AI
• MiniMax H3
• Text-to-video generation
• Pika H3 Flex

Topics:AI CompaniesGenerative AIVideo Generation

References

1. ^1 ^2 ^3 ^4 ^5 ^6 ^7Wiggers, K. "Pika, which is building AI tools to generate and edit videos, raises $55M." *TechCrunch*, 28 November 2023.
2. ^1 ^2 ^3 ^4 ^5 ^6 ^7 ^8 ^9 ^10 ^11 ^12 ^13 ^14 ^15 ^16 ^17 ^18 ^19 ^20 ^21 ^22 ^23 ^24 ^25 ^26Pika. "AI Company Pika Raises $55M to Redesign Video Making and Editing." *Business Wire* (press release), 28 November 2023.
3. ^1 ^2 ^3 ^4 ^5 ^6 ^7 ^8 ^9 ^10 ^11 ^12Guo, D.; Meng, C. "Pika raises $80M, so anyone can make video on command." Pika blog (archived copy), 5 June 2024.
4. ^1 ^2 ^3 ^4Bastian, M. "AI video generator Pika Labs raises 80 million US dollars." *The Decoder*, 5 June 2024.
5. ^1 ^2 ^3 ^4 ^5 ^6 ^7 ^8 ^9 ^10Goldman, S. "This 27-year-old's TikTok-like AI app makes playful, creative short videos from just a few words." *Fortune* (Eye on AI), 16 October 2025.
6. ^1 ^2 ^3 ^4Gray, J. "Generative Video Startup Pika Labs Launches Version 1.0, Raises $55 Million in Funding." *PetaPixel*, 28 November 2023.
7. ^1 ^2 ^3 ^4Pika. "Terms of Service." Mellis, Inc., last modified 4 August 2026.
8. ^1 ^2Franzen, C. "Pika Labs raises $55M, launches new AI video platform to take on Runway." *VentureBeat*, 28 November 2023.
9. ^1 ^2 ^3 ^4 ^5 ^6 ^7 ^8McKay, C. "Pika Labs Secures $80M in Series B Funding." *Maginative*, 7 June 2024.
10. ^1 ^2 ^3 ^4 ^5 ^6 ^7 ^8Lightspeed Venture Partners. "Pika's Imagination Engine." 28 November 2023.
11. ^1 ^2 ^3 ^4 ^5 ^6Angelo, J. "Video AI company Pika offers new tool to speed up AI video creation." *Semafor*, 17 September 2026.
12. ^1 ^2 ^3 ^4Pika (@pika_labs). "Introducing Pika AI Selves." X (Twitter), 20 February 2026.
13. ^1 ^2 ^3 ^4 ^5Pika. "Welcome to the Pika API Club." Pika blog, 5 August 2026.
14. ^1 ^2 ^3 ^4 ^5Pika. "Everything You Want to Hear: Introducing Pika Audio." Pika blog, 14 August 2026.
15. ^1 ^2 ^3 ^4 ^5 ^6Pika. "Welcome to the New Pika." Pika blog, 17 September 2026.
16. ^1 ^2 ^3 ^4Pika. "Pika Is Now Inside ChatGPT." Pika blog, 16 October 2026.
17. ^1 ^2 ^3 ^4 ^5 ^6 ^7Bao Lam; Phong Ngo. "Meet Demi Guo, Harvard graduate behind $470M AI video startup Pika." *VnExpress International*, 25 October 2025.
18. ^Rives, A.; Meier, J.; Sercu, T.; Goyal, S.; Lin, Z.; Liu, J.; Guo, D.; et al. "Biological structure and function emerge from scaling unsupervised learning to 250 million protein sequences." *Proceedings of the National Academy of Sciences* 118 (15), April 2021.
19. ^Guo, D.; Rush, A. M.; Kim, Y. "Parameter-Efficient Transfer Learning with Diff Pruning." ACL 2021. arXiv:2012.07463.
20. ^1 ^2 ^3Meng, C. "Chenlin Meng's Homepage." Accessed 27 October 2026.
21. ^Song, J.; Meng, C.; Ermon, S. "Denoising Diffusion Implicit Models." ICLR 2021. arXiv:2010.02502. Code: ermongroup/ddim.
22. ^Meng, C.; He, Y.; Song, Y.; Song, J.; Wu, J.; Zhu, J.-Y.; Ermon, S. "SDEdit: Guided Image Synthesis and Editing with Stochastic Differential Equations." ICLR 2022. arXiv:2108.01073.
23. ^Meng, C.; Rombach, R.; Gao, R.; Kingma, D. P.; Ermon, S.; Ho, J.; Salimans, T. "On Distillation of Guided Diffusion Models." CVPR 2023 (award candidate). arXiv:2210.03142.
24. ^1 ^2"Chenlin Meng." *100 Women in AI*. Accessed 27 October 2026.
25. ^"Pika Raises $35M in Series A Funding." *FinSMEs*, 28 November 2023.
26. ^"Pika." Lightspeed Venture Partners portfolio page. Accessed 27 October 2026.
27. ^"Pika valuation, funding & news." *Sacra*. Accessed 27 October 2026.
28. ^"Meta Platforms Accelerates AI Push Through Acquisitions, Talent Recruitment, and Heavy Investment." *Stock News* (via Tiger Brokers), 1 August 2025.
29. ^1 ^2 ^3"Pika's Latest AI Video Model Lets You Melt, Explode, and Cake-ify Objects." *PetaPixel*, 2 October 2024.
30. ^Bastian, M. "Pika Labs releases AI video generator 2.0 with new features." *The Decoder*, 14 December 2024.
31. ^1 ^2Pika (@pika_labs). "Pika 2.2 is HERE, with 10s generations, 1080p resolution, and Pikaframes." X (Twitter), 27 February 2025.
32. ^1 ^2 ^3Adobe. "Adobe Firefly Revolutionizes Creative Ideation with New Mobile App, Multimedia Moodboarding and Expanded AI Models." *Business Wire* (press release), 17 June 2025.
33. ^Pika. "Pika 2.5." pika.art model page. Accessed 27 October 2026.
34. ^1 ^2"Pika's Demi Guo launches 'AI Selves', a personalized always-on avatar that learns and monetizes your expertise." *TBPN Digest*, 26 February 2026.
35. ^1 ^2 ^3Pika. "Introducing Real-Time Video Chat for Agents." Pika blog, 2 April 2026.
36. ^1 ^2Pika. "Say Hello to Pika Speech." Pika blog, 18 August 2026.
37. ^1 ^2"Pika - AI Video." App Store listing (developer: Mellis, Inc.). Accessed 27 October 2026.
38. ^1 ^2Pika. "Your Pika Agent Just Got a Job." Pika blog, 10 April 2026.
39. ^Pika. "Every Breakthrough Starts With an Experiment." Pika blog, 26 May 2026.
40. ^Pika. "Get Your Product Out There." Pika blog, 28 May 2026.
41. ^"Pika revenue, valuation and employees." *GetLatka* (third-party estimates). Accessed 27 October 2026.
42. ^1 ^2 ^3Pika Team. "Introducing Pika H3 Flex." Pika blog, 29 October 2026.
43. ^1 ^2Pika. "Pika-Labs/Pika-H3-Flex." Hugging Face model card. Accessed 1 November 2026.

Show all 43 references