AIAcademy · on Fugu Ultra
The 5-hour limit has been exceeded, so I have to wait 4 hours. However, it kindly provides guidance … I like this one better because it is user-oriented, offering friendly guidance for beginners and general users.
anecdotal
In the wild
Field reports from real model users
Real posts from people using these models on actual prompts and tasks — praise and criticism alike, each linked to its source. Qualitative feedback, never scored or ranked.
Official benchmarks show reported scores. In-the-wild reports show what users hit after release: latency, cost, quota, regressions, surprising wins, and task-specific failures. These are source-linked anecdotes, not benchmark scores.
Models
1 selected
Fugu UltraSakana AI
Report tone
Report type
Topic
Mixed
AM9:21 · on Fugu Ultra
It discovered 27 bugs that Fable 5 couldn't find and fixed all of them. The code quality is impeccable … it implemented about 70,000 lines of new features, resolved 4 issues, and created 7 PRs.
Task Introduced Fugu into a repo previously worked on with Claude Fable 5; ~1 hour of use.
anecdotaloutput shown
am.will · on Fugu Ultra
The game was pretty bad and notably worse than GPT 5.5. … GPT 5.5 by contrast did a pretty good job and required no follow ups.
Task Asked it to build a Three.js replica of Rocket League via Codex.
anecdotalprompt shownsingle run
Mark Santos · on Fugu Ultra
In terms of model speed and performance, Fugu on Opencode won … inverted directional turn, wonky camera, no sfx, not identical to Crossy Road game.
Task Head-to-head vs Claude Opus 4.8: a single-file Three.js Crossy Road game.
prompt shownsingle run