Pelican Test: Opus 5.5 uses 2x tokens of Fable 5.2 for only 4% more cost — and makes mistakes
PilgrimofHaqq2 · reddit · 2026-11-11
A Reddit user ran the same "go all out" prompt through Fable 5.2 Low and Opus 5.5 Med, each producing a single-file HTML pelican animation.
- Fable 5.2 Low: 594 lines / 40KB / 14 min / 51,958 output tokens (45% thinking), $3.71 API-equivalent cost. Clean, no errors found — but minimal: one sunset scene, one click interaction, no sound.
- Opus 5.5 Med: 1,085 lines / 68KB / 29.5 min / 116,722 tokens (60% thinking), $3.87. Far more ambitious: self-synthesized sound effects, a full day/night cycle with lamps and fireflies, jumping fish on click, keyboard shortcuts — but visible bugs: two sets of wings, beak above the eyes, scarf clipping through the pelican.
Key takeaway: Opus used 2x the time and tokens for only 4% more cost, doing twice the work with flaws, while Fable did half the work correctly.
Related event: Pelican Test: Opus 5.5 Burns 2x Tokens for 4% Higher Cost(2 posts)→
More from Models
- OpenAI docs still mention "models 6.2 and later", hinting at unreleased GPT-6.2 — Angaisb_ · 2026-11-11
- Collaborator model Ampersand lands on Vercel AI Gateway, free for a week at 744B params — ericzelikman · 2026-11-11
- Same API price, 8x cheaper: Muse Ember builds voxel scene for $3.09 vs Haiku 5.5's $24.96 — alexandr_wang · 2026-11-11
- Anthropic's Fable 5.2 shows major robot-control gains, beats purpose-built systems on RoboDojo — DJiafei · 2026-11-11
- NVIDIA launches Alpamayo 2 Edge, first real-time Thor-ready reasoning model for vehicles — drmapavone · 2026-11-11
- OpenRouter officially joins Stripe, vowing to connect money rails with AI intelligence rails — jeff_weinstein · 2026-11-11