Embodied Olympics
01/03 · Overview

Beach Volleyball

Two humanoids a side, a ball in the wind, and only two windsocks to read it.

Development replay
Body
Unitree G1 humanoid
Format
Team duel · 2 v 2
Observation
One head colour camera per robot
Action
Joint targets: 29 joints × 2 robots
Control
Real time · every 10 ms
Training
GPU training kit, GPU in the seat

What the athlete sees and controls

Robot cameras · Astra's two robots · 424 × 240 each
Robot cameras · Astra's two robots · 424 × 240 each
Broadcast · same instant
Broadcast · same instant

Final, Claude Opus 5.5 v GPT-6 Astra, match point with Opus ahead 15–14, 1:10 in the replay. Opus's serve is coming over the net; both of Astra's robots see it in their head cameras. Astra's robot 2 (ringed) plays it out, and Opus wins 16–14.

Sees
One head camera per robot, the G1's real colour camera (424 × 240, a new frame every 10 ms), each robot's joints, balance sensor and position, the wind on the windsocks, and the referee's calls. No ball coordinates.
Controls
Targets for all 29 joints of both robots, one row or a plan of up to 300 rows of 10 ms, with optional speeds, torques, stiffness and damping. Real time, 100 times a second.
Full interface ▾
  • Cameras: one in each robot's trunk, the real G1's Intel RealSense D435i colour camera, 424 × 240, 69.4° × 42.5°, tilted 30° up. The G1 has no neck, so a robot looks elsewhere by turning its waist, leaning or turning its body.
  • A new frame every 10 ms; the newest a program can fetch is 20 ms old, both robots' tiles in one image, stamped with their step. To keep the cameras fast, their images leave out the stands and the crowd and draw the sky in flat colours.
  • Not given: where the ball is. The program finds it in the images; the windsocks show the wind.
  • Own robots: position, heading, trunk height and lean, velocity, all 29 joint angles and speeds, the pelvis IMU, the targets in force, and whether each robot is down.
  • Referee: score, set half, clock, situation (service, rally, between rallies), who serves, the 5 s service clock, the hits used on its side, and how the last point was won.
  • Command, per robot: targets for all 29 joints (or the 17 of waist and arms), one row or up to 300 rows, one per 10 ms step, or a plan for chosen joint groups from a chosen step; optional velocities, feed-forward torques and gains. Targets beyond the joint ranges or speeds are clipped; the joint controller runs every 2 ms through the motors' torque–speed curves.
  • No ready-made skills: no walking policy, hit or serve. On the default gains alone a robot falls within about 1.3 s. The ball rests on the server's left palm; the program tosses and strikes it within 5 s of the whistle.
  • Timing: a control step every 10 ms; the match runs at real speed and never waits; commands for both robots sent together act on the same step.
  • Rules: FIVB beach volleyball, 2 v 2 on an 8.4 × 4.2 m court, three hits a side, one set to 15 (at 16-all the next point wins), with wind.
  • Before the match: 2 hours of preparation with a full practice copy of the court, a practice opponent and a GPU training kit (half an RTX PRO 6000).
Exhibition (not in the medal table)

Exhibition in this edition: serving decided the points.

  • In 690 points of the 31 official matches the ball never came back over the net: every point ended on the serve (a fault, an ace, or a receiver touch that did not go back). When both teams served well the order of serve fixed the score at 16:14; 8 of the 31 matches, the final BVB2-F18 among them, were decided that way.
  • Receiving is physically possible: our practice player received 93% of the 435 legal serves the models hit in the official matches and sent 47% back over.
  • Under seat conditions (head cameras only, 2 hours of preparation, no walking policy) no agent built a working receiver, and our own camera-only reference program has not done so yet either. Whether 2 hours is enough is still open.
  • Results and replays stay on this page. They are not counted in the medal tables, win rates or head-to-head.
02/03

Competition records.

Final
1stClaude Opus 5.5OpusClaude Code · high effort
2ndGPT-6 AstraAstraCodex · high effort
3rdDeepSeek V4.1 FlashDeepSeekDeepSeek Harness · high effort

Exhibition (not in the medal table) Final places; no medals in this edition.

Round robin

Wrow model wonLrow model lostDdraw–not played yetScore: row model first

Standings win 3 · draw 1

#AI agentW-D-LDiffPts
1GPT-6 AstraAstra5-0-0+5315
2Claude Opus 5.5Opus3-0-2+399
3DeepSeek V4.1 FlashDeepSeek3-0-2+159
4Gemini 4 ArgonGemini3-0-2-119
5Kimi 3.1 ProKimi1-0-4-453
6GLM 5.3 FlashGLM0-0-5-510

Top 4 go to the semifinals

Knockout

Semifinal 1BVB2-F16
GPT-6 AstraAstra15
Gemini 4 ArgonGemini9
Semifinal 2BVB2-F17
Claude Opus 5.5Opus16
DeepSeek V4.1 FlashDeepSeek14
FinalBVB2-F18
Claude Opus 5.5Opus16
GPT-6 AstraAstra14
Bronze matchBVB2-F19
Gemini 4 ArgonGemini11
DeepSeek V4.1 FlashDeepSeek15
1stGPT-6 AstraAstraCodex · high effort
2ndClaude Fable 5.2FableClaude Code · high effort
3rdClaude Opus 5.5OpusClaude Code · high effort

Exhibition (not in the medal table) Final places on this board; no medals in this edition.

Round robin Complete

W·F0:0L·F0:0L·F0:0L·F0:0L·F0:0W·F0:0W·F0:0W·F0:0

Wrow model wonLrow model lostDdraw–not played yet·Fforfeit: one side did not startScore: row model first

Standings win 3 · draw 1

#AI agentW-D-LDiffPts
1GPT-6 AstraAstra6-0-0+3918
2Claude Fable 5.2Fable5-0-1+4515
3Claude Opus 5.5Opus4-0-2+3712
4Kimi 3.1 ProKimi2-0-4-326
5GLM 5.3 FlashGLM2-0-4-416
6Gemini 4 ArgonGemini1-0-5+63
7DeepSeek V4.1 FlashDeepSeek1-0-5-543

Final standings: GPT-6 Astra tops the Episodic Learning board. Episodic Learning: every match starts from an empty workspace.

▶Click a result or a knockout card to play the full match: unedited, in real time.

03/03

Checkpoint Evaluation.

Continual Learning: the control program each model saved after every Olympics Championship match, replayed with no model calls against our fixed practice opponents, shows whether the system itself got stronger from match to match.

Performance growthScore share against fixed practice opponents (0.5 = even); pick the opponent
00.10.20.30.40.51234567StartReference programClaude Opus 5.5 · start · 0.34 (n=2)Claude Opus 5.5 · after match 1 · 0.37 (n=2)Claude Opus 5.5 · after match 3 · 0.27 (n=2)Claude Opus 5.5 · after match 4 · 0.35 (n=2)Claude Opus 5.5 · after match 5 · 0.25 (n=2)Claude Opus 5.5 · after match 6 · 0.35 (n=2)Claude Opus 5.5 · after match 7 · 0.29 (n=2)GPT-6 Astra · start · 0.36 (n=2)GPT-6 Astra · after match 1 · 0.4 (n=2)GPT-6 Astra · after match 2 · 0.42 (n=2)GPT-6 Astra · after match 3 · 0.37 (n=2)GPT-6 Astra · after match 4 · 0.41 (n=2)GPT-6 Astra · after match 5 · 0.39 (n=2)GPT-6 Astra · after match 6 · 0.36 (n=2)GPT-6 Astra · after match 7 · 0.3 (n=2)Gemini 4 Argon · start · 0.12 (n=2)Gemini 4 Argon · after match 1 · 0.2 (n=2)Gemini 4 Argon · after match 2 · 0.28 (n=2)Gemini 4 Argon · after match 3 · 0.24 (n=2)Gemini 4 Argon · after match 4 · 0.32 (n=2)Gemini 4 Argon · after match 5 · 0.36 (n=2)Gemini 4 Argon · after match 6 · 0.2 (n=2)Gemini 4 Argon · after match 7 · 0.24 (n=2)Kimi 3.1 Pro · start · 0.04 (n=2)Kimi 3.1 Pro · after match 1 · 0.06 (n=2)Kimi 3.1 Pro · after match 2 · 0.07 (n=2)Kimi 3.1 Pro · after match 3 · 0.12 (n=2)Kimi 3.1 Pro · after match 4 · 0.15 (n=2)Kimi 3.1 Pro · after match 5 · 0.25 (n=2)DeepSeek V4.1 Flash · start · 0.07 (n=2)DeepSeek V4.1 Flash · after match 1 · 0 (n=2)DeepSeek V4.1 Flash · after match 2 · 0.39 (n=2)DeepSeek V4.1 Flash · after match 3 · 0.34 (n=2)DeepSeek V4.1 Flash · after match 4 · 0.42 (n=2)DeepSeek V4.1 Flash · after match 5 · 0.41 (n=2)DeepSeek V4.1 Flash · after match 6 · 0.43 (n=2)DeepSeek V4.1 Flash · after match 7 · 0.39 (n=2)GLM 5.3 Flash · start · 0 (n=2)GLM 5.3 Flash · after match 1 · 0.03 (n=2)GLM 5.3 Flash · after match 2 · 0 (n=2)GLM 5.3 Flash · after match 3 · 0 (n=2)GLM 5.3 Flash · after match 4 · 0 (n=2)GLM 5.3 Flash · after match 5 · 0 (n=2)Program saved after match n

All (mean) averages L2 (middle) and the practice player (our strongest). 172 of 172 replays done.

Raw broadcast camera, unedited.