Frontier LLM scores on real robot benchmark double every 6 weeks, 15% to 48% in 3 months
ycombinator · x · 2026-10-14
Robocurve's real-robot benchmark (RoboDojo-RC Tier 1, 120 attempts per model, 95% CIs) shows frontier LLM mean task scores climbing exponentially: the record went from 15% (GPT-5.6 Sol) in July 2026 to 48% (Opus 5.5 XH) by September — roughly doubling every 6 weeks. The chart tracks GPT-6 Astra, Gemini 3.8 Flash, Opus 5/5.5, and GPT-6/6.1 Sol against release dates, suggesting rapid transfer of model capability gains into real robot tasks.
More from Embodied
- Recova Trains Robots to Self-Recover in Simulation, Cutting Human Intervention From 87.5% to 0% — Lianhuiq · 2026-10-14
- NVIDIA's AlphaNIST Learns Robotic Assembly From Scratch With Only Binary Rewards, Hitting 98.2% Across 20 Tasks — abhishekunique7 · 2026-10-14
- EU privacy watchdog calls for smart glasses to clearly signal when recording — Polymarket · 2026-10-14
- How Argon drives Spot: real2sim skill verification in MuJoCo and map-free navigation — ki_ki_ki1 · 2026-10-14
- Agent Argon reads Spot's SDK docs itself, builds trash-sorting skill on Boston Dynamics robot — chris_j_paxton · 2026-10-14
- Meta researcher recruiting multimodal and robotics interns for next summer, no background needed — liuzhuang1234 · 2026-10-14