Benchmark Almanac

Embodied First Exam

GPT-6 Astra 32.7% leads since Oct 2026 Not in model cards

Coding agents answer frozen robot tasks from many benchmark suites with a program. Each program is verified in a separate sandbox, run on Harbor.

Released 2026-10-19. About 220 GitHub stars in its first weeks.

Made by Northwestern University and collaborators (CMU , UC Berkeley , Stanford and others) Released Oct 2026 Simulated robots All tasks public Last model added Sep 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
Code A program or patch run against tests.
mll-lab-nu/embodied-first-exam 224 stars Leaderboard embodied-first-exam.ai
7
Models scored on its own leaderboard

GPT-6 Astra leads with 32.7%

Best score per model from the benchmark's own leaderboard, read on 2 Nov 2026. Where a model ran with several settings or harnesses, we keep its best.

  1. 32.7 GPT-6 Astra
  2. 23.5 Claude Opus 5.5
  3. 19.7 Gemini 4 Argon
  4. 6.14 Gemini 3.8 Flash
  5. 3.95 Qwen 3.8 Max
  6. 2.63 GLM-5.3-Flash
  7. 2.63 DeepSeek V4.1 Flash
  1. 1 GPT-6 Astra high 32.7%
  2. 2 Claude Opus 5.5 high 23.5%
  3. 3 Gemini 4 Argon high 19.7%
  4. 4 Gemini 3.8 Flash high 6.14%
  5. 5 Qwen 3.8 Max high 3.95%
  6. 6 GLM-5.3-Flash high 2.63%
  7. 7 DeepSeek V4.1 Flash high 2.63%
#ModelScoreSettingDateSource
1 GPT-6 Astra 32.7% high, Codex Sep 2026 Official leaderboard
2 Claude Opus 5.5 23.5% high, Claude Code Sep 2026 Official leaderboard
3 Gemini 4 Argon 19.7% high, Antigravity Sep 2026 Official leaderboard
4 Gemini 3.8 Flash 6.14% high, Antigravity Sep 2026 Official leaderboard
5 Qwen 3.8 Max 3.95% high, Claude Code Aug 2026 Official leaderboard
6 GLM-5.3-Flash 2.63% high, Claude Code Aug 2026 Official leaderboard
7 DeepSeek V4.1 Flash 2.63% high, dsh Sep 2026 Official leaderboard

Number one over time

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

OpenAI
Own leaderboard 1 leader
GPT-6 Astra 32.7%

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% Near the ceiling (85%) GPT-6 Astra Gemini 3.8 Flash 0% 20% 40% 60% 80% 100% Near the ceiling (85%) GPT-6 Astra Gemini 3.8 Flash

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%