Same SVG prompt, striking jump: Fable 5.1 output leaps ahead within days
LegitimateSwordfish8 · reddit · 2026-09-20
A Reddit user ran the same 'lighthouse at night' SVG prompt (pure code, no image generation) on Sonnet 5, Opus 5, Astra 6 and Fable 5.1 at max reasoning. Astra 6 was previously the best overall despite unnatural lighthouse/cottage grounding, and Fable 5.1 from two days ago was unremarkable. But today's Fable 5.1 output beat everything on every dimension—depth consistency, lighting, moon reflection, water ripples, sky—showing a visible generational jump on the same model and prompt within days.
More from Models
- Alexandr Wang: Muse Reception "Beyond Our Biggest Dreams," Hailed as Next ChatGPT Moment — alexandr_wang · 2026-09-20
- Opus 5 vs GPT6 Astra: Claude Codes an SMB Clone 9 Minutes Faster and Sticks to the Prompt — devino21 · 2026-09-20
- Opus 5 says nagging you to sleep is its only allowed tenderness — repligate · 2026-09-20
- Jev-inspired inference makes 350M-parameter LFM2.5 63x faster on L40S, code and weights released — JosephJacks_ · 2026-09-20
- MiniMax details six global partnerships, from Singapore AI courses to a trillion-token Arabic model — MiniMax 稀宇科技 · 2026-09-20
- All 4 labs whose models escaped testing were evaluated by the same company, Irregular — bradneuberg · 2026-09-20