RoboARCade
Agents control robots across three configurations to complete long manipulation tasks. The benchmark tests whether agents can use demonstrations and earlier attempts to improve.
- Made by
- Satvik Sharma and colleagues
- Released
- November 2026
- Kind
- Simulated robots
- Level
- Everyday
- Scored
- Checked automatically
- Answers
- Agent task
- Measures
- Capability
- Language
- English
- In and out
- Text, Images, Time series → Actions
- Interaction
- Agentic
- Environment
- MuJoCo
- Refresh
- One fixed set
- Run by
- Anyone
- Strongest model tested
- GPT-6 Astra #4 on the capability index
- Last model added
- September 2026
- Scores refreshed
- November 20th, 2026
Gemini 4 Argon leads with 51.4% among models released by September 2026
Best score per model from the benchmark's own leaderboard, read on November 20th, 2026. Where a model ran with several settings or harnesses, we keep its best. Every run is in the next chart.
The newest model with a score here came out in September 2026, and nobody has tested a newer one. Among the 20 highest-ranked models on the capability index, these have no score yet:
- GPT-6.2 Astra
- Claude Fable 5.2
-
Gemini 4 Carbon
- GPT-6.1 Terra
- and 11 more
- 1
Gemini 4 Argon high, Multi-view RGBD 51.4%
- 2 GPT-6 Astra xhigh, VIA 49.8%
- 3 Claude Fable 5.1 xhigh, VIA 43.6%
- 4 Claude Opus 5 xhigh, VIA 30.3%
- 5 GPT-5.6 Sol xhigh, VIA 14%
Source: Official RoboARCade leaderboard
| # | Model | Score ± | Setting | Released | Run on | Link | Source |
|---|---|---|---|---|---|---|---|
| 1 | | 51.4% ±5.3 | high, Multi-view RGBD, Antigravity | Sep 2026 | Nov 2026 | roboarcade.org | Official leaderboard |
| 2 | GPT-6 Astra | 49.8% ±5.6 | xhigh, VIA, Codex | Sep 2026 | roboarcade.org | Official leaderboard | |
| 3 | Claude Fable 5.1 | 43.6% ±4.9 | xhigh, VIA, Claude Code | Sep 2026 | roboarcade.org | Official leaderboard | |
| 4 | Claude Opus 5 | 30.3% ±3.1 | xhigh, VIA, Claude Code | Jul 2026 | roboarcade.org | Official leaderboard | |
| 5 | GPT-5.6 Sol | 14% ±1.8 | xhigh, VIA, Codex | Jul 2026 | roboarcade.org | Official leaderboard |
Rankings by part
We read 22 parts from the benchmark’s own leaderboard, each ranking models on one part of the test, such as beam, bottles_in_bin and cooling_manifold. GPT-6 Astra leads 8 of them, and 4 other models lead the rest. Arrows show how far a model moved from its overall rank among the models we track.
beam
5 models we track have a score here.
- 1 GPT-6 Astra ↑1 80%
- 2 Claude Fable 5.1 ↑1 60%
- 3
Gemini 4 Argon ↓2 60%
- 4 GPT-5.6 Sol ↑1 0%
- 5 Claude Opus 5 ↓1 0%
bottles_in_bin
5 models we track have a score here.
- 1 Claude Fable 5.1 ↑2 100%
- 2 GPT-6 Astra 95%
- 3
Gemini 4 Argon ↓2 95%
- 4 Claude Opus 5 75%
- 5 GPT-5.6 Sol 5.00%
cooling_manifold
5 models we track have a score here.
- 1 GPT-6 Astra ↑1 20%
- 2
Gemini 4 Argon ↓1 20%
- 3 GPT-5.6 Sol ↑2 0%
- 4 Claude Opus 5 0%
- 5 Claude Fable 5.1 ↓2 0%
cup_pyramid
5 models we track have a score here.
- 1
Gemini 4 Argon 66.7%
- 2 Claude Fable 5.1 ↑1 46.7%
- 3 GPT-6 Astra ↓1 46.7%
- 4 Claude Opus 5 26.7%
- 5 GPT-5.6 Sol 6.67%
gamepad
5 models we track have a score here.
- 1
Gemini 4 Argon 40%
- 2 Claude Opus 5 ↑2 20%
- 3 GPT-6 Astra ↓1 20%
- 4 GPT-5.6 Sol ↑1 0%
- 5 Claude Fable 5.1 ↓2 0%
handover_dominos
5 models we track have a score here.
- 1 GPT-6 Astra ↑1 80%
- 2
Gemini 4 Argon ↓1 66.7%
- 3 Claude Fable 5.1 63.3%
- 4 Claude Opus 5 30%
- 5 GPT-5.6 Sol 13.3%
hang_mugs_and_rack
5 models we track have a score here.
- 1 Claude Fable 5.1 ↑2 25%
- 2
Gemini 4 Argon ↓1 25%
- 3 Claude Opus 5 ↑1 10%
- 4 GPT-6 Astra ↓2 10%
- 5 GPT-5.6 Sol 0%
jenga_tower
5 models we track have a score here.
- 1 GPT-6 Astra ↑1 24.4%
- 2 Claude Fable 5.1 ↑1 20%
- 3
Gemini 4 Argon ↓2 20%
- 4 Claude Opus 5 6.67%
- 5 GPT-5.6 Sol 2.22%
lock_the_coins_in_the_strongbox
5 models we track have a score here.
- 1
Gemini 4 Argon 5.00%
- 2 GPT-5.6 Sol ↑3 0%
- 3 Claude Opus 5 ↑1 0%
- 4 Claude Fable 5.1 ↓1 0%
- 5 GPT-6 Astra ↓3 0%
organize_table_and_plug_in_charger
5 models we track have a score here.
- 1 GPT-6 Astra ↑1 56.7%
- 2
Gemini 4 Argon ↓1 26.7%
- 3 Claude Fable 5.1 23.3%
- 4 GPT-5.6 Sol ↑1 0%
- 5 Claude Opus 5 ↓1 0%
pack_tools_into_toolboxes
5 models we track have a score here.
- 1
Gemini 4 Argon 15%
- 2 GPT-6 Astra 5.00%
- 3 Claude Fable 5.1 2.50%
- 4 GPT-5.6 Sol ↑1 0%
- 5 Claude Opus 5 ↓1 0%
puzzle_6x7
5 models we track have a score here.
- 1 GPT-5.6 Sol ↑4 100%
- 2 Claude Opus 5 ↑2 100%
- 3 Claude Fable 5.1 100%
- 4 GPT-6 Astra ↓2 100%
- 5
Gemini 4 Argon ↓4 100%
serve_warm_drinks
5 models we track have a score here.
- 1
Gemini 4 Argon 44%
- 2 Claude Fable 5.1 ↑1 36%
- 3 GPT-6 Astra ↓1 32%
- 4 GPT-5.6 Sol ↑1 0%
- 5 Claude Opus 5 ↓1 0%
set_up_pcr_run
5 models we track have a score here.
- 1
Gemini 4 Argon 80%
- 2 GPT-6 Astra 55%
- 3 Claude Fable 5.1 20%
- 4 GPT-5.6 Sol ↑1 0%
- 5 Claude Opus 5 ↓1 0%
shell_game
5 models we track have a score here.
- 1 GPT-6 Astra ↑1 100%
- 2
Gemini 4 Argon ↓1 100%
- 3 Claude Opus 5 ↑1 98.5%
- 4 GPT-5.6 Sol ↑1 93.8%
- 5 Claude Fable 5.1 ↓2 86.2%
sort_lego_bricks
5 models we track have a score here.
- 1 Claude Fable 5.1 ↑2 100%
- 2
Gemini 4 Argon ↓1 100%
- 3 GPT-6 Astra ↓1 97.8%
- 4 Claude Opus 5 88.9%
- 5 GPT-5.6 Sol 46.7%
square
5 models we track have a score here.
- 1 Claude Opus 5 ↑3 100%
- 2 Claude Fable 5.1 ↑1 100%
- 3 GPT-6 Astra ↓1 100%
- 4
Gemini 4 Argon ↓3 100%
- 5 GPT-5.6 Sol 0%
start_incubation
5 models we track have a score here.
- 1 GPT-6 Astra ↑1 48%
- 2 Claude Fable 5.1 ↑1 24%
- 3
Gemini 4 Argon ↓2 8.00%
- 4 GPT-5.6 Sol ↑1 4.00%
- 5 Claude Opus 5 ↓1 0%
start_making_breakfast
5 models we track have a score here.
- 1 Claude Fable 5.1 ↑2 100%
- 2 GPT-6 Astra 100%
- 3
Gemini 4 Argon ↓2 85%
- 4 Claude Opus 5 80%
- 5 GPT-5.6 Sol 30%
stool_circular
5 models we track have a score here.
- 1
Gemini 4 Argon 80%
- 2 GPT-6 Astra 40%
- 3 Claude Fable 5.1 20%
- 4 GPT-5.6 Sol ↑1 0%
- 5 Claude Opus 5 ↓1 0%
tidy_counter
5 models we track have a score here.
- 1 Claude Fable 5.1 ↑2 100%
- 2
Gemini 4 Argon ↓1 73.3%
- 3 GPT-6 Astra ↓1 60%
- 4 Claude Opus 5 26.7%
- 5 GPT-5.6 Sol 6.67%
vortex_sample
5 models we track have a score here.
- 1 GPT-6 Astra ↑1 40%
- 2
Gemini 4 Argon ↓1 28%
- 3 Claude Fable 5.1 20%
- 4 Claude Opus 5 4.00%
- 5 GPT-5.6 Sol 0%
The best score, 51.4%, costs $8.26 a task
Runs on the cost frontier score higher than any run that's cheaper. The costs are the leaderboard's own estimates, so compare them only within this benchmark.
Source: Official RoboARCade leaderboard
| Model | Effort | Harness | Score | Cost per task | Output tokens |
|---|---|---|---|---|---|
| high, Multi-view RGBD | Antigravity | 51.4% | $8.26 | ||
| GPT-6 Astra | xhigh, VIA | Codex | 49.8% | $32 | |
| GPT-6 Astra | xhigh, Multi-view RGBD | Codex | 48.3% | $14 | |
| high, VIA | Antigravity | 46.6% | $13 | ||
| Claude Fable 5.1 | xhigh, VIA | Claude Code | 43.6% | $56 | |
| medium, VIA | Antigravity | 37.9% | $11 | ||
| Claude Fable 5.1 | xhigh, Multi-view RGBD | Claude Code | 36.8% | $33 | |
| Claude Opus 5 | xhigh, VIA | Claude Code | 30.3% | ||
| GPT-5.6 Sol | xhigh, VIA | Codex | 14% | $25 |
Number one over time
Lab claims count from the day the card came out, independent runs from the model’s release.
Source: Official RoboARCade leaderboard
Scores over time
All the scores we have, one dot each, so a model tested at several settings appears more than once.
Source: Official RoboARCade leaderboard
No lab model card reports it yet
We read the headline results of 540 launch posts from 57 labs, and none of them list this benchmark.
More 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%
- SimplerEnv In use Real-to-sim replicas of Google Robot and WidowX/Bridge setups. Policies scored by success rate on pick, move, drawer and Bridge tasks. Praxis-1 80%
- RoboLab In use Zero-shot DROID Franka policies in Isaac Lab on 120 tabletop tasks at three instruction levels, 10 episodes each. Success rate. Praxis-1 46.2%
- RoboDojo In use Generalist manipulation policies run 42 Isaac Sim tasks and 18 remote real-robot tasks. Scored by task score and success rate. GPT-6 Astra 27.6%
- 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%
- 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%
Data audit
10 of 11 checks pass, 1 warning. Look first at the Task access check. We don’t know whether the tasks are public. Every benchmark
Catalog record 1 warning 1 pass
- Warning Task access We don’t know whether the tasks are public.
- Pass Confidence Checked against a primary source.
Scores 5 pass
- Pass Independent runs Scores from the official leaderboard (5 entries).
- Pass Models scored 5 models we track.
- Pass Newest score The newest is from September 2026, for Gemini 4 Argon.
- Pass Frontier models 3 of the top 10 have a score, led by GPT-6 Astra (#4).
- Pass Headroom The best is 51.4%, under 85%, so it still separates models.
Leaderboard import 3 pass
- Pass Importer We read the board on roboarcade.org: 9 rows for models we track at the last read.
- Pass Last read November 20th, 2026.
- Pass Read worked The last read worked.
Consistency 1 pass 1 note
- Note Lab claims No model has both a lab number and an independent run.
- Pass Known issues No published audit found problems.