Benchmark Almanac

RoboColiseum

GO-3 Nano Preview 78.8% tops its leaderboard Claude Opus 5.5 76.3% leads the models we track Not in model cards

AgiBot Genie robot in Genie Sim 3.0 (Isaac Sim): instruction, robustness, manipulation and spatial boards after post-training on provided sim data.

Instruction board crowded (many pi0.5 fine-tunes 0.72-0.84). Manipulation and spatial not saturated. Many noisy third-party rows. Organizer baselines mixed with submissions.

Made by AgiBot Released Jan 2026 Simulated robots Last model added Oct 2026 Scores refreshed 2 Nov 2026
Level
Everyday Most adults, without training
Scored
Checked automatically A program checks the result against an answer key, unit tests or the expected final state.
Answers
Agent task Many steps in a terminal, browser, desktop, app, game or robot.
AgibotTech/genie_sim 1,472 stars Paper Leaderboard
25
Entries on its own leaderboard, 9 models we track

GO-3 Nano Preview tops the leaderboard with 78.8%

Best score per entry from the benchmark's own leaderboard, read on 2 Nov 2026. Where a model ran with several settings or harnesses, we keep its best. Names in gray aren't in our model list: fine-tunes, agent systems, products and older models. They don't count toward leaders. The best model we track is Claude Opus 5.5, at 76.3%.

25 entries. Scroll sideways for the rest.

  1. 78.8 GO-3 Nano Preview
  2. 77.5 Paimon
  3. 76.9 HiDream-O1-Embodied
  4. 76.3 Claude Opus 5.5
  5. 76.0 GPT-6 Astra
  6. 70.9 GPT-6.1 Sol
  7. 70.8 DM0.5
  8. 66.9 Gemini 4 Argon
  9. 65.0 GigaBrain-0.7
  10. 59.8 WNDFC
  11. 59.4 Xiaomi-Robotics-1
  12. 58.4 OpenWAM-Alpha
  13. 57.9 pi0.5_refine
  14. 57.4 π0.5
  15. 56.7 kp-robot
  16. 56.3 ACoT-VLA
  17. 55.8 lingbot-vla-v2
  18. 55.1 DeepSeek V4.1 Flash
  19. 47.2 Dream-1 SFT 500 steps
  20. 46.8 GR00T N1.7
  21. 46.7 InternVLA-A1.5
  22. 43.0 Xiaomi-Robotics-0
  23. 42.3 Muse Wonder
  24. 33.5 Dream-1 zero-shot
  25. 28.9 π0
  1. 1 GO-3 Nano Preview 78.8%
  2. 2 Paimon 77.5%
  3. 3 HiDream-O1-Embodied 76.9%
  4. 4 Claude Opus 5.5 76.3%
  5. 5 GPT-6 Astra 76%
  6. 6 GPT-6.1 Sol 70.9%
  7. 7 DM0.5 70.8%
  8. 8 Gemini 4 Argon 66.9%
  9. 9 GigaBrain-0.7 65%
  10. 10 WNDFC 59.8%
  11. 11 Xiaomi-Robotics-1 59.4%
  12. 12 OpenWAM-Alpha 58.4%
  13. 13 pi0.5_refine 57.9%
  14. 14 π0.5 57.4%
  15. 15 kp-robot 56.7%
  16. 16 ACoT-VLA 56.3%
  17. 17 lingbot-vla-v2 55.8%
  18. 18 DeepSeek V4.1 Flash 55.1%
  19. 19 Dream-1 SFT 500 steps 47.2%
  20. 20 GR00T N1.7 46.8%
  21. 21 InternVLA-A1.5 46.7%
  22. 22 Xiaomi-Robotics-0 43%
  23. 23 Muse Wonder 42.3%
  24. 24 Dream-1 zero-shot 33.5%
  25. 25 π0 28.9%
#EntryScoreSettingDateSource
1 GO-3 Nano Preview 78.8% Oct 2026 Official leaderboard
2 Paimon 77.5% Oct 2026 Official leaderboard
3 HiDream-O1-Embodied 76.9% Sep 2026 Official leaderboard
4 Claude Opus 5.5 76.3% Sep 2026 Official leaderboard
5 GPT-6 Astra 76% Sep 2026 Official leaderboard
6 GPT-6.1 Sol 70.9% Sep 2026 Official leaderboard
7 DM0.5 70.8% Sep 2026 Official leaderboard
8 Gemini 4 Argon 66.9% Sep 2026 Official leaderboard
9 GigaBrain-0.7 65% Aug 2026 Official leaderboard
10 WNDFC 59.8% Aug 2026 Official leaderboard
11 Xiaomi-Robotics-1 59.4% Oct 2026 Official leaderboard
12 OpenWAM-Alpha 58.4% Oct 2026 Official leaderboard
13 pi0.5_refine 57.9% Jul 2026 Official leaderboard
14 π0.5 57.4% Apr 2025 Official leaderboard
15 kp-robot 56.7% Jul 2026 Official leaderboard
16 ACoT-VLA 56.3% Jun 2026 Official leaderboard
17 lingbot-vla-v2 55.8% Oct 2026 Official leaderboard
18 DeepSeek V4.1 Flash 55.1% Sep 2026 Official leaderboard
19 Dream-1 SFT 500 steps 47.2% Oct 2026 Official leaderboard
20 GR00T N1.7 46.8% Apr 2026 Official leaderboard
21 InternVLA-A1.5 46.7% Sep 2026 Official leaderboard
22 Xiaomi-Robotics-0 43% Aug 2026 Official leaderboard
23 Muse Wonder 42.3% Oct 2026 Official leaderboard
24 Dream-1 zero-shot 33.5% Oct 2026 Official leaderboard
25 π0 28.9% Oct 2024 Official leaderboard

3 models have held number one since its release in Jan 2026

Lab claims count from the day the card came out, independent runs from the model’s release.

Physical Intelligence OpenAI Anthropic
Own leaderboard 3 leaders
Claude Opus 5.5 76.3%

Scores over time

All the scores we have, one dot each, so a model tested at several settings appears more than once.

On the benchmark’s own leaderboard Best so far
0% 20% 40% 60% 80% 100% 2025 2026 Near the ceiling (85%) Claude Opus 5.5 π0.5 π0 0% 20% 40% 60% 80% 100% 2025 2026 Near the ceiling (85%) Claude Opus 5.5 π0.5

No lab model card reports it yet

We read the headline results of 479 launch posts from 51 labs, and none of them list this benchmark.

More in Robotics

All 86 in Robotics
  • ERQA In use 400 multiple-choice visual questions from robot and egocentric footage on spatial, trajectory, action and state reasoning. Accuracy. Gemini Robotics ER 2 78.5%
  • RoboLab In use Zero-shot DROID Franka policies in Isaac Lab on 120 tabletop tasks at three instruction levels, 10 episodes each. Success rate. GR00T N1.7 8.58%
  • Human proximity (Google DeepMind robotics) In use Models judge from robot camera views whether a human is within 1 meter, a spatial awareness check for safe robot behavior. Scored by accuracy. Gemini Robotics ER 2 93%
  • Moment finding (Google DeepMind) In use Models find the moment in a robot video when an event happens. Scored by accuracy and mean absolute time distance in seconds. Gemini Robotics ER 2 91.3%
  • Progress classification (Google DeepMind) In use Models assign each frame of a robot video to one of five task-progress bands from 0 to 100 percent. Scored by accuracy. Gemini Robotics ER 2 57.4%
  • LIBERO Fading Simulated Franka tabletop tasks in four 10-task suites. Policies fine-tuned on 50 demos per task. Success rate over ~50 trials. MolmoAct 2 97.2%