Speak Language Learning App Secures $20M Funding, Doubling Its Valuation

AI-powered language learning app Speak is experiencing remarkable growth. Since its launch in South Korea in 2019, Speak has amassed over 10 million users, according to CEO and co-founder Connor Zwick. The app's user base has doubled annually for the past five years, and it now serves customers in more than 40 countries.

In light of this impressive expansion, investors are stepping up to support the startup, pledging new funding. This week, Speak announced the closure of a $20 million Series B extension, led by Buckley Ventures, with contributions from the OpenAI Startup Fund, Khosla Ventures, Paul Graham, co-founder of Y Combinator, and LinkedIn executive chairman Jeff Weiner. This funding round elevates Speak's total capital raised to $84 million and doubles the startup's valuation to $500 million.

Founded in 2014 by Zwick and Andrew Hsu, who met while participating in the Thiel Fellowship, Speak aims to teach languages through speaking practice and repetition, rather than rote memorization of vocabulary and grammar. This approach bears some resemblance to Duolingo, especially with its recent generative AI features. However, True to its namesake, Speak prioritizes verbalizing as the core component of language learning.

“Our core philosophy is all about encouraging users to speak out loud as much as possible,” Zwick explained. “Achieving fluency enables people to forge connections, bridge cultures, and create economic opportunities. Speaking remains the most crucial aspect of language acquisition, yet it has historically been the least supported by technology.”

Initially focusing on English, Speak has since expanded to include Spanish lessons, which utilize a speech recognition model developed from proprietary data. French lessons are next on the horizon, although Zwick did not specify a launch date.

The app operates on a subscription model, charging $20 per month or $99 annually for comprehensive access to its features, including review materials and specialized courses. With a 75-person team across offices in San Francisco, Seoul, Tokyo, and Ljubljana (the capital of Slovenia), Speak's near- to long-term plans involve developing new models to enhance real-time feedback on tone and pronunciation, Zwick noted.