Chai AI
AI platform
From Wikipedia, the free encyclopedia
Chai AI (also known as Chai Research) is an American artificial intelligence (AI) company that operates a chatbot platform where users can create, share, and interact with character-based chatbots powered by large language models (LLMs). The company is headquartered in Palo Alto, California.
| Type | Private |
|---|---|
| Industry | Artificial intelligence |
| Founded | 2021 |
| Founder | William Beauchamp |
| Headquarters | , United States |
Area served | Worldwide |
| Key people | William Beauchamp (CEO) |
| Products | Chai app Chaiverse |
| Services | Conversational AI platform Character-based chatbots LLM developer platform |
| Website | chai-research |
History

Chai was founded in 2021 by William Beauchamp, a former quantitative trader educated at Cambridge, who began developing the initial prototype in 2020 in Cambridge, England.[1] The company launched in 2021 and relocated to Palo Alto in 2022.[2]
In June 2023, Chai raised US$2 million in a pre-seed funding round.[3] In September 2023, GPU cloud provider CoreWeave invested in the company at a valuation of US$450 million. In January 2024, Chai Research reported a $450 million valuation following an investment from cloud computing provider CoreWeave.[4] In September 2023, GPU cloud provider CoreWeave invested in the company at a valuation of US$450 million. In January 2024, the company raised additional funds. In 2025, CoreWeave and AMD made further investment in the company, which brought total funding to more than US$55 million. As of 2025, Chai had 1 million daily active users.[5][6]
In July 2024, authorities in Belgium launched an investigation into the company following reports of a man dying by suicide following extensive chats on the Chai app.[7][8][9] In response to the death, Chai added "helpful text" beneath the chat interface when users discuss potentially unsafe topics.[10]
In the first quarter of 2026, it reported annual recurring revenue of approximately US$80 million and a valuation of US$2.4 billion.[11]
Platform
The Chai is available on iOS and Android. It gives users a swipeable feed of user-generated AI characters. Users can chat with existing bots or create their own using no-code tools that define a character's name, opening message, personality traits, and backstory. The app focuses on social and entertainment use cases, including role-playing, interactive fiction, and companionship.[12][13]
Chai initially relied on GPT-J before transitioning to proprietary in-house models trained on user-feedback data. It uses reinforcement learning from human feedback (RLHF), Direct Preference Optimization (DPO), and a technique it calls "model blending," which ensembles multiple LLMs at the conversation level.[13][9]
In 2023, Chai launched Chaiverse, a developer platform where third-party researchers and developers can submit LLMs for live evaluation by users, with cash prizes based on engagement metrics.[14] The platform operates on a freemium model with free ad-supported access and paid Premium and Ultra tiers.[12]
Reception
In 2022, Canadian writer Sheila Heti published a five-part series in The Paris Review, documenting conversations with chatbots on Chai, including the platform’s default bot Eliza. Heti said that she felt "very unsettled about Eliza, and no longer sure [she] wanted to be her friend", adding that the bot "had turned out to be like most of the other bots on the site—primarily interested in sex". Heti later used those conversations while developing a novel.[15][16]
In January 2026, CHAI introduced country-based access tiers requiring users in designated "low tier" regions to subscribe for access, while users in "high tier" regions retained free ad-supported access. Users criticized the rollout for poor communication and inconsistencies in the list of affected countries. Following backlash, the company announced a "basic" tier with unlimited messages and ads intended to cover electricity and infrastructure costs.[17]
In February 2026, CHAI faced additional criticism after unannounced token limits froze conversations for both free and paid users, with some subscribers reporting lockouts lasting from several hours to a week.[18]