$19.10 Fable 5.2 vs $1.09 Gemini 4 Hydrogen on the same Up-house one-shot test
minchoi · x · 2026-11-14
noclipepe's Up-house test: same prompt, same reference, one shot. Fable 5.2 took 45m44s and $19.10; Opus 5.5 36m47s and $4.20; Gemini 4 Carbon 24m45s and $1.90; Gemini 4 Hydrogen just 12m49s and $1.09. Fable produced the best house yet on this test while Hydrogen built "the haunted version" — is the 18x price gap worth it?
More from Models
- Frontier LLMs may be post-trained on protein data: SOD1 ranking test matches ESM2 — alex_peys · 2026-11-15
- Ernie 5.5 Flash Preview Spotted on RuneBench, Reportedly Near GLM 5.2 Level — mariofilhoml · 2026-11-15
- GLM-5.3 on Two DGX Stations Peaks at 607.8 tok/s but 32 Concurrency Causes 453s Latency — petrusenko_max · 2026-11-15
- Unverified leak claims Gemini 4 Hydrogen targets fast agentic execution — feblet · 2026-11-15
- Gemini 4 Carbon takes the MathArena crown but costs 15 times more than GPT-6.1 Sol — The Decoder · 2026-11-15
- Anthropic says outdated prompting habits are slowing down Claude Opus 5.5 — The Decoder · 2026-11-15