AI for superhumans
Farang's next-generation AI architecture is designed to help you solve the hardest problems and to excel at tasks where the best alternatives underperform. It is designed not just for today, but for a future world where AI is a casual part of daily life.
Going beyond current limitations
Large language models (LLMs) have revolutionized the world in almost every field, ranging from basic personal assistance to medical diagnosis and software development. But the architectures powering those successes were not originally designed to do so and seems to have succeeded by chance. We believe in a designed approach, inspired by the biological brain and standing on the learnings from existing LLMs.
Farang, based in Sweden, was founded in 2025 from a scientific breakthrough which was the result of two years of research. The company has secured initial funding and is now scaling up the technology to compete with the current market leaders.
Research frontiers
Progress requires a dual focus: architecting the long-term future while advancing the present. Our research operates on two parallel tracks to achieve this.
Foundational Research: FX Architecture
Our primary mission is developing the FX architecture, a new foundation for AI designed from first principles for superior reasoning and memory. This is our long-term project to move beyond the limitations of today's models.
Open Science & Contributions
Alongside our core research, we are committed to advancing the entire field. We will actively publish findings and release open-source projects that complement and challenge the current state-of-the-art in AI.
Get Early Access
Be among the first to experience the next generation of AI. Join our waitlist for exclusive updates and early access to our models.