No Claude Fable 5? No problem: Sakana achieves frontier performance with new Fugu multi-model, auto synthesis system
June 22, 2026
Last night, the increasingly enterprise-focused AI startup Sakana launched Fugu, a multi-agent orchestration system that delivers frontier-level AI performance through a single, OpenAI-compatible API.
Designed for developers, enterprises, and nations seeking resilience against vendor lock-in and geopolitical export controls, Fugu (Japanese for “pufferfish”), bypasses the traditional monolithic model structure by dynamically routing queries to a swappable pool of specialized AI agents.
Sakana CEO and co-founder David Ha, formerly of Google Brain, positioned Fugu as a more reliable option for enterprise workflows than any single AI model provider in the wake of Anthropic’s move on June 12 to revoke public access to its most powerful models, Claude Mythos 5 and Claude Fable 5, in the wake of a U.S. government export control order.
“Fugu dynamically orchestrates the world’s best models to tackle complex tasks. We are proving that a well-orchestrated pool of swappable agents can match restricted frontier models like Fable and Mythos. But Fugu is about more than just performance. I believe that Orchestration Models are the next frontier, beyond bigger models. Relying on a single company’s model for national infrastructure is a massive risk. As recent export controls have shown, access to top models can disappear overnight. Collective intelligence is the practical hedge against this concentration of power. Fugu simply routes around vendor restrictions by relying on an entirely swappable agent pool.”
Sakana AI explicitly states that the specific models Fugu selects and how it coordinates them are proprietary, meaning this routing information is hidden from the user by design.
How Sakana Fugu works
At its core, Sakana Fugu operates like a master general contractor. When presented with a complex request, Fugu does not attempt to execute every step itself. Instead, it breaks the problem down, delegates sub-tasks to a pool of expert foundation models, verifies their work, and synthesizes the final output.
“Fugu is itself an LLM, trained to call various LLMs in an agent pool, including instances of itself recursively,” the Sakana AI team noted in their technical release.
Grounded in two of Sakana’s 2026 research papers, TRINITY and the Conductor, the system autonomously manages the entire lifecycle of model selection and verification using learned coordination strategies rather than hand-designed workflows.
Product tiers
- Fugu: A high-speed, low-latency model optimized for everyday tasks and coding environments like Codex.
- Fugu Ultra: The flagship tier engineered for complex, high-stakes tasks such as AI research, cybersecurity analysis, and multi-step patent investigations.
Benchmark performance
- LiveCodeBench: Fugu Ultra (93.2), Fugu (92.9) vs Claude Fable 5 (89.8)
- GPQA-D: Fugu Ultra (95.5) vs Claude Mythos Preview (94.6)
- SWE-Bench Pro: Fugu Ultra (73.7) vs Claude Opus 4.8 (69.2) and GPT-5.5 (58.6)
- CTI-REALM cybersecurity: Fugu Ultra (69.4) vs Opus 4.8 (69.6)
Pricing and availability
Fugu Ultra pricing: 30 per million output tokens. Subscription tiers: Standard 100/month, Max $200/month.
Fugu is available immediately in most regions, with temporary restriction from the EU and EEA while Sakana works to align its black-box data routing architecture with GDPR regulations.
Sakana AI was formed in Tokyo in 2023 by Llion Jones (co-author of “Attention Is All You Need”) and David Ha. The company reached a $2.6 billion Series B valuation in late 2025.