TLDR: FriendliAI, a leader in AI inference technology, has successfully closed a $20 million seed extension funding round. The capital will be used to expand its platform, which aims to significantly reduce the high costs and slow speeds associated with deploying large-scale AI models. The company’s technology has demonstrated the ability to cut GPU costs by up to 90% and deliver market-leading performance for Large Language Models (LLMs).
FriendliAI, a pioneering company in artificial intelligence (AI) inference technology, announced today the successful completion of a $20 million seed extension funding round. This significant investment, led by Capstone Partners and joined by Sierra Ventures, Alumni Ventures, KDB, and KB Securities, is earmarked for the expansion of FriendliAI’s innovative platform. The platform is specifically designed to address the critical challenges of high infrastructure costs and slow inference speeds that hinder the widespread deployment of large-scale AI models in real-world applications.
As businesses increasingly transition their AI projects from experimental stages to practical, real-world applications, they frequently encounter substantial obstacles. These include escalating infrastructure expenses and sluggish inference speeds, which can severely impact user experience and operational efficiency. FriendliAI directly tackles these issues by offering a platform that has been proven to reduce GPU costs by an impressive ‘up to 90%’ while simultaneously providing ‘market-leading performance for Large Language Models (LLMs)’.
Byung-Gon Chun, CEO and founder of FriendliAI, emphasized the growing demand for efficient AI solutions, stating, ‘The transition of AI from an experimental technology to a core business tool has created a surge in demand for efficient and scalable inference solutions.’ He further added, ‘This funding is a strong validation of our approach. We are thrilled to partner with our investors to scale our platform and help more companies unlock the full potential of AI without the prohibitive costs.’
FriendliAI’s advanced technology has already garnered considerable recognition and adoption within the industry. The company recently forged a strategic partnership with LG AI Research, becoming the ‘exclusive provider of API access to LG’s advanced EXAONE 4.0 foundation model’. Furthermore, several prominent AI companies are already leveraging the FriendliAI platform to achieve substantial gains. For instance, Scatter Lab, a popular character AI chatbot company in Korea, has ‘drastically reduced its infrastructure costs for its Zeta LLM service’. Similarly, Upstage, renowned for its cutting-edge Solar foundation model, relies on FriendliAI to ensure ‘state-of-the-art inference performance’.
Investors have expressed strong confidence in FriendliAI’s technical prowess and strategic market positioning. Eun-gang Song, Partner at Capstone Partners, remarked, ‘FriendliAI’s ability to deliver superior performance while significantly reducing costs makes them an essential partner for any enterprise scaling its AI operations.’ Ben Yu, Partner at Sierra Ventures, echoed this sentiment, adding, ‘FriendliAI is providing the critical infrastructure needed to power the next wave of AI adoption.’
Also Read:
- Farang Secures €1.5 Million Seed Funding to Revolutionize AI with Novel Foundational Architecture
- US AI Startups Secure Over $100 Million in Funding: 33 Companies Achieve Mega-Rounds in 2025
The newly secured capital will be strategically allocated to bolster FriendliAI’s sales and marketing initiatives, scale its operations to effectively meet the escalating customer demand, and continue its relentless innovation in platform development. FriendliAI is also noted as the ‘only provider supporting over 430,000 models on Hugging Face’, offering extensive model coverage for developers. Access to Friendli Inference is available through FriendliAI’s Dedicated Endpoints, Serverless Endpoints, and Container solutions.
