Gemini 4 Carbon tops Blueprint-Bench 2 physical-world benchmark, Claude Fable 5.2 second
cgarciae88 · x · 2026-11-12
On Blueprint-Bench 2, an AndonLabs benchmark where AI agents draw floorplans from photographs of apartment interiors, Google's Gemini 4 Carbon now ranks #1, edging out Anthropic's Claude Fable 5.2 at #2. Both models reportedly understand the physical 3D world almost as well as humans, per the benchmark's framing.
More from Models
- GPT-Live talk at VapiCon: OpenAI pushes natural speech with frontier intelligence — juberti · 2026-11-12
- VIALS-AA leaderboard goes live: compare models on score, cost and tool calls — ArtificialAnlys · 2026-11-12
- Tool use cost overhead varies wildly: Qwen3.8 Max pays 28× per task, GPT-6 models barely 1.3× — ArtificialAnlys · 2026-11-12
- Tool use nearly doubles VIALS-AA scores: tool-less Claude Fable 5.2 loses to Claude Haiku 5.5 with tools — ArtificialAnlys · 2026-11-12
- ChatGPT refuses to draw "sexy Chinese ants," user memes it with $NVDA — generativist · 2026-11-12
- Developers say persistent agents run on Anthropic while OpenAI lags on memory — charlespacker · 2026-11-12