Is LMArena quietly testing a Gemini 4 stealth model? SVG quality sparks suspicion
Background-Youth7289 · reddit · 2026-09-20
A Reddit user compared anonymous LMArena models on an SVG 'horse riding a bike' prompt. The model labeled Gemini 3.8 Flash High took 30 minutes to render vs 5 minutes for Claude Fable 5.1 High — opposite of earlier reports — and produced dramatically better output than expected from a Flash-tier model. The author suspects LMArena was quietly testing an experimental, possibly Gemini 4-generation model under that label, though it's unverified speculation. Worth watching for corroborating reports.
More from Models
- Datalab claims parsing accuracy second only to Gemini at 1/37 the cost and 6x speed — VikParuchuri · 2026-11-12
- Datalab fast-mode OCR beats Opus 5.5 and GPT-6.1 Sol at 1/4 to 1/68 the price — VikParuchuri · 2026-11-12
- Gemini 4 Carbon opens to all at $2/$10, price doubling to $4/$20 on Dec 1 — socialwithaayan · 2026-11-12
- Ten wild Gemini 4 Carbon demos in 24 hours: CUDA meadows, V8 engines, mech stomps — socialwithaayan · 2026-11-12
- How to cut Fable 5.2 token costs by 90%: Arc routing plus a lean CLAUDE.md — EXM7777 · 2026-11-12
- Grok Bot 1.0 launches with support for joining Meet, Zoom and Teams calls — mark_k · 2026-11-12