'Draw 24 with exactly 24 circles' test: Gemini fails, Opus 5.5 and ChatGPT pass
davidmanheim · x · 2026-10-11
In a widely shared test, models were asked to draw the number 24 using exactly 24 circles. Researcher David Manheim confirms Gemini failed even on high thinking, calling it a second-tier model, while Opus 5.5 on Medium and ChatGPT on High had no problem. The prompt: white background, black circles, exactly 24 circles, followed by asking the model how many it drew.
More from Models
- Claude Models on Google Cloud (Anthropic managed) is now generally available — claudeai · 2026-10-23
- Beam vs GLM 5.3 Abliterated: 10-round attack/defense test shows 96% of vulns patched — bigblueboo · 2026-10-23
- ChatGPT web can actually listen to audio, but the same model can't via API or Codex — Possible_Statement84 · 2026-10-23
- MiniMax M3.1 Flash free for 3 days on 0G, with 1M context and 512K output — MiniMax_AI · 2026-10-23
- TLA+ expert corrects Grok: TLC only exhausts finite state spaces, not all behaviors — tianyin_xu · 2026-10-23
- Open LLM cost calculator shows Arabic tokenization and 6x price gaps across models — Moham-saed · 2026-10-23