ModelFugu Ultra v2 by Sakana AI
Text+image+file->text
Free credits included. No credit card required.
Fugu Ultra v2 is the highest-capability model in Sakana AI's Fugu family, designed for complex multi-step reasoning, autonomous research, and full-stack software development. Rather than a single monolithic model, Fugu is a learned multi-agent orchestration system — a language model trained to route tasks across a swappable pool of open-weight and specialized models, including NVIDIA's Nemotron family, and to recursively call instances of itself. Fugu Ultra v2 excels at tasks requiring sustained reasoning over complex visual and structured data, and achieves frontier-level performance without relying on any single proprietary model, providing resilience against vendor lock-in. It supports a 1M token context window and multimodal inputs including text, images, and files.
