Qualia Beats Flora and GPT-6 Astra Across Three Robot-Data Benchmarks
lukas_m_ziegler · x · 2026-10-29
Fabian shares evaluation results: Qualia scores very high on three important public robot-data benchmarks, scored exactly the way NVIDIA scores them—his setup reproduces NVIDIA's reported numbers for GPT-6 Astra and Flora within about a point (side note: Flora is surprisingly good).
Tasks include cutting robot video into steps, describing them, ranking recordings, and judging how far a robot got—Qualia leads Flora and GPT-6 Astra on every one. The author says they're just getting started and expects Qualia to improve significantly in the coming days and weeks, and hopes more people will try it.
More from Models
- Researcher: Codex $500 plan burns credits faster than Anthropic's $20 sub — sethlazar · 2026-10-29
- Model trained via Tinker beats rival on finance benchmark at 9% of the cost — simonguozirui · 2026-10-29
- Uncensored GGUF Quant of LiquidAI's LFM3-1.7B Trends on Hugging Face — AtomicChat · 2026-10-29
- Claude Pro vs ChatGPT Plus at $20: robotics dev crowdsources which holds up — Wild_Shower6854 · 2026-10-29
- Odyssey-3 world models now simulate general environments in real time with higher physical fidelity — Scobleizer · 2026-10-29
- Particle-1 inference goes live on OpenRouter, claiming 28-52% fewer output tokens — ycombinator · 2026-10-29