Leaker claims GPT-6 Astra is a 4.2T-param, 120B-active looped model; Opus 5.5 may not be looped
scaling01 · x · 2026-09-28
Leaker scaling01 says he's holding back his methodology article for further analysis, but shares his latest central estimates: GPT-6 Astra is likely a 4.2T total-parameter model with 120B active, around 112 layers, looping 50% of its layers once. He also claims evidence suggests GPT-6 Sol, Luna, and Opus 5.5 are not looped language models. Leaks about "Astra Minor" point to a smaller Sol-sized looped model that likely slots into the $20-25 price tier. All figures are unconfirmed speculation.
Related event: Leaker estimates GPT-6 'Astra' as 4.2T-parameter looped language model(5 posts)→
More from Models
- TypeSafe AI's Jev returns confidence-scored decisions, not text: a16z sits down with Diogo Almeida — a16z · 2026-09-28
- Meta recaps 6 months of launches: 10 Muse releases plus Meta Model API — AIatMeta · 2026-09-28
- Vercel CEO: open models now drive 80% of Vercel's token traffic as OpenAI, Gemini, Anthropic lose ground — RihardJarc · 2026-09-28
- Real-time robot control showdown: GPT-6 Astra vs Opus 5.5 vs Grok 4.7 vs MolmoAct2 — DJiafei · 2026-09-28
- Open Models Now 4 Months Behind Frontier, Handle 56% of Vercel Gateway Tokens — bigdata · 2026-09-28
- Users Report Being Routed to Unreleased Claude Sonnet 5.5, Even on Free Tier — lyraxana · 2026-09-28