MazeBench Recap: Astra With Code Execution Cuts Planning Time From 20+ Minutes to Under 5
patience_cave · x · 2026-09-14
A repost of MazeBench numbers for GPT-6 Astra: with Python, 23% after 400M tokens and 24 hours of reasoning (14% without). With tools it clears intro levels easily, thinking under 5 minutes per attempt versus a 9-minute average (peaks above 20) without — a 60% gain at half the cost via its world model.
Related event: GPT-6 Astra on MazeBench: Bare Model Beats GPT-5.6 With Tools(6 posts)→
More from Models
- Anthropic removes 5-hour usage cap for verified Claude Ultra subscribers — james_mtc · 2026-10-31
- DeepSeek's silent downtime upgrade caught by the pelican bicycle test — teortaxesTex · 2026-10-31
- ARC Prize 2026 closes in 3 days as Yi-Chia Chen tops ARC-AGI-3 at 75.56% — GregKamradt · 2026-10-31
- Fable 5.5 Max stuns with gorgeous SVG of a pelican riding a bicycle, headed to Claude Code — ChrisGPT · 2026-10-31
- OpenAI reportedly to train on users' Daybreak sessions with no opt-out — dyn___ · 2026-10-31
- NVIDIA Nemotron 3.5 scales LLM self-verification, +9-10 points on agentic benchmarks — sarahcat21 · 2026-10-31