Kimi K3.1 Pro Matches Top Tier on Robot Benchmarks at 20x Cost
On the BiGym 2 humanoid robot benchmark, Kimi K3.1 Pro scored 21.3%, placing it in the top tier alongside Gemini and GPT models but at roughly 20x the cost of GPT-6 Sol. Per-task analysis suggests the top two labs have already been training on robot-related tasks.
2026-10-26 ~ 2026-10-26 · 3 related posts
- Kimi K3.1 Pro scores 21.3% on humanoid BiGym 2 benchmark at 20x GPT-6 Sol's cost — teortaxesTex · 2026-10-26
- Robotics eval breakdown: Kimi K3.1 lands between frontier models but costs 20x more — teortaxesTex · 2026-10-26
- Per-task robotics eval breakdown suggests top 2 labs already training on robot tasks — teortaxesTex · 2026-10-26