- Chinese cybersecurity firm 360 reportedly unveiled Tulongfeng, claiming it can go head-to-head with Anthropic‘s Mythos.
- Tokyo-based Sakana AI launched Fugu, which it says “stands shoulder-to-shoulder with leading models like Anthropic’s Fable 5 and Mythos Preview.”
- Both arrive as the US government’s export ban on Anthropic’s Mythos and Fable 5 — issued about two weeks ago — drags on.
- Sakana markets Fugu as delivering “frontier capability without the risk of export controls.”
What Happened
Two Asian AI startups have launched frontier models pitched directly against Anthropic’s export-controlled systems, TechCrunch reports. Chinese cybersecurity firm 360 unveiled Tulongfeng, which it says can go head-to-head with Anthropic’s Mythos. Earlier the same week, Tokyo-based Sakana AI launched Fugu — named for the Japanese blowfish — claiming it “stands shoulder-to-shoulder with leading models like Anthropic’s Fable 5 and Mythos Preview.”
Why It Matters
The launches are a direct market response to the US order that pulled Mythos and Fable 5 from non-American users roughly two weeks ago. Export controls intended to keep frontier capability inside the US appear to be spurring rivals to fill the vacuum — the classic unintended consequence of restriction, and a fresh front in the US-China AI race.
Technical Details
Tulongfeng targets the same cybersecurity niche as Mythos, the model so capable the Trump administration banned it (and the more restricted Fable 5) for non-Americans. Fugu is designed for agents, with the ability to orchestrate access to other models through their APIs. Sakana — which recently opened a recursive self-improvement lab — said Fugu “is something we have been building since last year,” with research presented at ICLR this spring.
Who’s Affected
Anthropic now faces unrestricted rivals for the exact capability the US fenced off. Non-US enterprises locked out of Mythos and Fable 5 gain export-control-free alternatives. US policymakers face evidence the ban may be backfiring: Sakana’s site explicitly advertises “frontier capability without the risk of export controls,” and a spokesperson called the timing “entirely coincidental” while conceding it drew attention.
What’s Next
The open question is capability: both companies claim parity with Anthropic’s models, but those claims are unverified and the banned Mythos line already pushed the limits of independent benchmarks. If the rivals prove genuinely competitive, the export ban will have accelerated a parallel non-US frontier-AI ecosystem rather than contained one.
