Sakana AI has announced the launch of Sakana Fugu, a next-generation multi-agent AI orchestration platform designed to help organizations tackle complex coding, scientific reasoning, research, and business intelligence tasks through a single API interface. The new platform marks a significant step in the company’s vision of advancing collective intelligence by enabling multiple AI models to work together as a coordinated system.
Unlike conventional AI solutions that rely on a single foundation model, Sakana Fugu dynamically orchestrates a network of specialized large language models, intelligently selecting, coordinating, and validating outputs to deliver more effective results for complex, multi-step workflows. The platform is designed to simplify AI adoption by allowing users to access advanced multi-agent capabilities through a unified endpoint without managing the complexity of multiple underlying models.
Built on Sakana AI’s research into collective intelligence and agent-based systems, Fugu leverages a lightweight orchestration model that learns how to route tasks across AI systems, enabling adaptive problem-solving and scalable reasoning. The platform has already been utilized internally by Sakana AI’s research and engineering teams and is now being made available to external users through an API-based offering and beta access program.
Also Read: Digital Envoy Launches LocID Native App on Snowflake to Enable Privacy-First Identity Collaboration with Zero Data Egress
Sakana AI stated that Fugu is engineered to support a broad range of enterprise and research use cases, including software development, scientific analysis, cybersecurity, knowledge discovery, and advanced reasoning workflows. By combining multiple frontier AI models into a coordinated framework, the company aims to provide organizations with greater flexibility, performance, and resilience while reducing dependence on a single AI provider.
The launch further strengthens Sakana AI’s growing portfolio of AI technologies and reinforces its commitment to developing innovative AI systems that combine collaboration, adaptability, and scalability to address increasingly complex challenges across industries.
SOURCE: Sakana