Leaks claim Google's new 'Krypton' model is stunning, hints of RSI-level self-improvement
kimmonismus · x · 2026-11-09
Multiple leaks are circulating about Google's latest model, codenamed "Krypton", with posters calling its outputs breathtaking. kimmonismus says he can't verify the leaks himself but argues Google is either heading toward or has already reached recursive self-improvement internally. Unconfirmed rumor, but the buzz suggests a major comeback for Google.
More from Models
- InferenceBench: agents beat default inference engines but mostly lose to hyperparameter search — maksym_andr · 2026-11-09
- InferenceBench trend: frontier models' inference speedups grew from 3x to 12x in a year — maksym_andr · 2026-11-09
- ChatGPT Pro $100 user burns just 1% of weekly limit despite heavy Codex and cloud agent use — filiksyos · 2026-11-09
- Power user: Opus 5.5 running 5 parallel sessions burned only 34% of weekly quota — kimmonismus · 2026-11-09
- InstaDeep CEO maps frontier progress onto AI2027: unreleased models halfway to Agents-2 — HankYeomans · 2026-11-09
- One prompt, pure code: Google model generates anime in 39 min at 105 tok/s — CurieuxExplorer · 2026-11-09