The graphs - every benchmark, drawn.
Every 0-10 the judge has handed out, for every model on the bench. Tap models to compare up to three - same tasks, same rubric, bars side by side. Claude Fable 5 leads off.
The deep dives — one full breakdown per model
official + independent benchmark tables, judge quotes, head-to-head records - Claude Fable 5 first
Claude Fable 58.10Fusion8.59Hermes MoA8.17GPT-5.6 Sol8.16LongCat-2.08.12Grok8.09MiniMax M37.97Fugu Ultra7.94Kimi K37.89GLM-5.27.77Fugu Mini7.75Opus 4.87.51Kimi K2.77.46Claude Sonnet 57.01Qwen 3.77.00Hy36.76Inkling6.07Kimi K2.7 · FastunrankedKimi K2.7 · No-ThinkunrankedKimi K2.7 · QualityunrankedClaude Mythos 5unrankedKilo Codeunranked
Category averages
Every task, bar by bar
grouped by category - hover a bar for the score - click a row for that task's page with every model's build
The same stack Julian uses
Run this stack yourself.
Every demo on this bench was built inside the Agent Operating System — one prompt, one shot, single HTML file out. The Agent OS, the prompts, the templates, the weekly walkthroughs and 4,000+ founders shipping with it every day all live inside the AI Profit Boardroom.
4,000+founders
258documented wins
38countries
$59/momonthly