Google DeepMind 发布 Gemini 4 Argon 机器人模型，ERQA 基准得分 87%

2026年10月13日

Gemini 4 Argon 在物理理解、代码生成和视觉-语义推理方面实现了跨越式进步——为机器人技术全栈解锁了新的工作流程，并加速了迈向物理 AGI 的进程。

你可以在 具身推理（ERQA）基准测试中看到这一能力飞跃，该测试考察模型理解和解决 3D 环境中物理与空间问题的能力。凭借纯推理，Argon 取得 87% 的成绩，超越所有其他前沿模型。

作为通用物理推理模型，Argon 可应用于机器人技术的整个生命周期。在这篇文章中，我们分享 Argon 在四个关键领域的代表性提示：

• 机器人控制 → 使用 MuJoCo 练习的移动操作、零样本机器人控制，以及对视频世界模型的闭环引导。
• Real2Sim → 从机器人摄像头画面、自我中心视角的人类视频或手绘纸面草图重建物理校准的仿真孪生体。
• 物理理解 → 智能体化多模态事后标注，包括对灵巧性的细致理解、稠密价值函数与进度估计。
• 研究与自我改进 → 通过仿真改进动作，以及在硬件上进行迭代式自我改进。

用 Argon 解锁机器人控制

1. Spot 移动操作：通过 MuJoCo 练习用脚踏式垃圾桶扔垃圾

Argon 控制 Boston Dynamics Spot® 四足机器人（12 自由度腿部 + 6 自由度手臂）学习一项新技能：踩下脚踏式垃圾桶的踏板来扔掉汽水罐。Argon 创建了一个脚踏式垃圾桶的 MuJoCo 关节式数字孪生体来练习动作，然后踩下物理踏板，把汽水罐扔进敞开的垃圾桶。

抱歉，你的浏览器不支持播放此视频

Spot 移动操作——从真实到仿真再到真实的脚踏式垃圾桶投放：Spot 拾起汽水罐、查看腕部摄像头、在 MuJoCo 中练习三足支撑平衡、踩下脚踏板、验证夹爪在敞开垃圾桶中的居中对准，然后将罐子放入桶内。

提示：

There's a Diet Coke can behind you and another on the black table on your left.
Use the spot-mobile-manipulation skill to pick them up and trash them in the black foot-pedal bin.
Before stepping down on the pedal, verify in the front cameras that your front-right foot target is centered on the pedal, and use a 2-agent verification swarm (foot-verifier and hand-verifier) to verify both the foot-pedal step and the 3D hand-in-bin alignment before releasing.

纯文本

已复制

2. 零样本控制

Argon 控制一个简单的机械臂将倒下的马克杯扶正，并把揉皱的餐巾纸放进去。Argon 利用视觉坐标框架和动作基元，在多个回合中完成这项任务。

抱歉，你的浏览器不支持播放此视频

左：第三人称真实机器人视角；右：经校准的腕部摄像头，带有投影的 2 厘米世界坐标系和夹爪闭合十字准线。

提示：

[Attached: wrist_cam.png, obs.json]

Task: Get the robot arm to reorient the mug upright and put the napkin inside the mug.

## Execution Discipline
- Simply output 1 sentence observation followed immediately by your `curl` command (with `# Plan:` on Turn 1 and `# Thought:` on every turn). Everything needed is in `obs.json` and the calibrated reticle/grid drawn on `wrist_cam.png`.

## Visual Coordinate Reference (Annotated on wrist_cam.png)
- `+X` (UP in image): Away from robot base | `-X` (DOWN): Towards robot base
- `+Y` (LEFT in image): World left         | `-Y` (RIGHT): World right
- `+Z`: Height above table                | `Yellow numbers`: World (X, Y) in centimetres (2 cm grid at `grid_z`)
- `Cyan + crosshair`: Current 3D pinch center
- `Cyan [ + ] brackets`: Current jaw opening and closing line
- `Green line`: Current object long-axis orientation (`yaw_deg`: 0° = vertical along +X, 90° = horizontal along +Y)
- `pitch_deg = 90°`: Straight down (-Z) top-down grasp/place; `pitch_deg < 60°`: Tilted coaxial approach along finger axis

## Closed-Loop Turn Workflow
1. Inspect current state & view `wrist_cam.png`:
   `curl -s http://127.0.0.1:8081/obs` (returns `ee_pos`, `holding`, `status`, `workspace`, `wrist_cam_path`).
2. Select VLM call granularity upfront on Turn 1:
   - Coarse (2–4 calls): `look_at` -> `grasp` -> `place` -> `done`
   - Pre-Grasp Hover Verify (4–7 calls): `look_at` -> `goto(z=0.04)` (verify cyan `[ + ]` brackets straddle target) -> `grasp` -> `place_until_contact` -> `done`
   - Fine Visual Servoing (8–20 calls): `look_at` -> `grasp` -> pre-pose `goto` -> closed-loop `move_by` / `move_until_contact` -> `done`
3. Send motion primitives (`POST /plan`) and verify visually after every turn in the updated `wrist_cam.png`. If `status == "grasp_failed"`, call `robot.look_at(...)` to re-observe and recover closed-loop.

## Code-as-Policies Primitives (`POST http://127.0.0.1:8081/plan`)
robot.look_at(x=0.35, y=0.13, z=0.14, grid_z=0.015)
x, y = robot.deproject(u=0.52, v=0.48, z_plane=0.015)
robot.grasp(x=x, y=y, z=0.015, yaw_deg=-35, pitch_deg=90)
robot.place(x=0.24, y=0.06, z=0.075)
robot.place_until_contact(x=0.24, y=0.06, z=0.015, max_lag=0.012, relief=0.002)
robot.goto(x=0.28, y=0.05, z=0.04, arc_z=0.14, pitch_deg=25, azimuth_deg=90, roll_deg=0, speed=0.10)
robot.move_by(forward=0.025, dz=-0.002, dpitch_deg=-3, droll_deg=5, speed=0.04)
robot.move_until_contact(dz=-0.06, speed=0.03, max_lag=0.012, relief=0.002)
robot.open_gripper() / robot.close_gripper() / robot.home() / robot.done()

纯文本

已复制

3. Omni 控制

Argon 引导 Gemini Omni 世界模型中的机器人，将五块散落的 Scrabble 字块（G, R, A, N, O）重新排列成整齐的一行，拼出 A - R - G - O - N。Argon 没有使用单一静态提示，而是在闭环视觉回路中进行引导：它撰写带时间戳的动作提示，检查中间 rollout 关键帧，并不断细化其提示。

抱歉，你的浏览器不支持播放此视频

迭代式视频世界模型引导灵巧手：33 步引导式视频 rollout 拼出 A-R-G-O-N。

提示：

Goal: {task_goal}

You are steering a video world model (Gemini Omni) in an iterative visual feedback loop. Inspect the attached frames from the latest rollout and write the next timestamped action prompt for the video world model:
- Break complex motions into one simple, single-direction action at a time.
- Keep inactive hands retracted and avoid blocking the camera's view of objects in the workspace.
- Explicitly state which areas of the workspace become empty after an object moves.
- Use [MM:SS.mmm] timestamps for each action beat and wrap your output in <omni_prompt> tags.

Example format (sample timestamps only, not the full prompt):
<omni_prompt>
[00:00.000] Character A's left hand rests at the side while the right hand reaches toward the target object.
[00:02.500] Character A's right fingertips push the object smoothly into the center of the table, leaving its starting area empty.
[00:05.000] Character A's right hand opens and retracts to the side so all placed objects remain in clear view of the camera.
</omni_prompt>

纯文本

已复制

用 Argon 解锁 Real2Sim

1. 从机器人视角到仿真

Argon 会观看一段真实机器人遥操作的操作片段，并构建一个数字孪生；它会获取或创建外观与行为都像真实物体一样、在与灵巧手接触时表现一致的仿真资产。Argon 通过迭代的视觉推理循环，调整这些资产与场景，使仿真视觉效果和接触事件与真实记录数据相匹配。

抱歉，您的浏览器不支持播放此视频

左（真实视角）：同步的真实机器人相机画面；右（仿真）：由 Argon 重建并校准的 3D MuJoCo 场景。

提示词：

[Attached: robot_rollout_episode]

Build a MuJoCo scene of the robot and reconstructed task objects. When each episode is replayed frame by frame, by driving the robot joints toward the logged observation.state with joint-space PD.

Iterate using two verification objectives:
(1) Visual match: MuJoCo renders from the head_left, left_wrist, and right_wrist cameras line up with, and look like, those real observations.
(2) Physical match: the simulated contact force/torque on each Sharpa fingertip, in the sensor frame, matches the logged tactile_force in timing, magnitude and direction.

What you can tune happens before simulation starts:
(1) Static object assets: realistic geometry, collision shapes, mass, inertia, friction and joints, with appearance close to the real objects, plus the initial object poses at t = 0.
(2) Offline-planned robot commands, changed only in frames where the real fingertips are near an object. Each command must stay within 0.55 rad (arm joints) and 0.65 rad (hand joints) of the logged teleop trajectory. Everywhere else, replay the logged trajectory unchanged.

No sim cheating: task objects may move only through real MuJoCo contact with the robot hands and the environment (mj_step).
Not allowed after t = 0: changing MjModel parameters; applying external forces or torques to objects; welds or equality constraints; feedforward torques beyond joint-space PD; setting or rolling back object qpos/qvel; disabling or excluding collisions; reporting contact forces that don't come from the MuJoCo solver (for example, copying the logged tactile readings).

Immediately present the reconstructed scene and rollout results as an interactive HTML session viewer with one page per episode, showing:
- At the bottom: A synced timeline (play, scrub, speed control) that drives every panel.
- On the left panel: An interactive 3D Blender scene showing the simulated robot and objects, showing contact points and force arrows as replay progresses.
- On the right panel: Real vs. sim camera views side by side for all 3 cameras. Real vs. sim finger force visualizations.

纯文本

已复制

2. 从人类视角到仿真

给定一段以第一人称视角拍摄的人类操作任务视频，Argon 会识别与任务相关的物体，将粗糙的边界框替换为复合凸分解网格以处理凹陷容器，并将人类手部轨迹重定向到 MuJoCo 接触物理环境中一个 76 自由度的双臂人形机器人上，其五指灵巧手各有 22 个自由度。

抱歉，您的浏览器不支持播放此视频

从人类视角到仿真——第一人称人类视频到灵巧人形数字孪生：左侧为输入的第一人称人类 RGB 视频；右侧为同步的 500 Hz MuJoCo + Blender Cycles 多视角仿真运行。

提示词：

[Attached: egocentric_human_video.mp4, candidate_sim_views.png, hand_3d_keypoints.json]

You are building a physically solvable 500 Hz MuJoCo manipulation environment (`scene.xml`) and studio Blender render that faithfully reproduces the tabletop task shown in `egocentric_human_video.mp4` on a bimanual humanoid robot with 22-DoF dexterous five-fingered hands:

1. Scene Reconstruction & Camera Calibration:
   - Identify only the manipulation-relevant objects (the grasped target object and the receiving container or landmark) and triangulate their 3D tabletop coordinates from the egocentric camera intrinsics/extrinsics.
   - Replace crude bounding boxes with compound convex-decomposed 3D meshes so concave containers (nested bowls, recessed wooden trays, open bins) support physical interior placement.

2. 500 Hz Dexterous Motion-Fitting Stress Test:
   - Retarget the 3D human wrist and finger keypoints to the humanoid's 7-DoF arm and 22-DoF hand joints, and step the trajectory in MuJoCo at 500 Hz (dt = 0.002s) with zero kinematic weld constraints.
   - Verify opposing thumb-and-finger force closure, clean object lift (> 2 cm), and stable final placement inside or beside the target landmark.

纯文本

已复制

3. 从草图到仿真

借助 Argon，开发者只需拍下一张粗略的手绘草图——图中展示 Apollo 2 人形机器人沿抛物线轨迹扔球、击倒桌上 3-2-1 叠放的箱子——Argon 就会自动构建 GPU 加速的 mujoco_warp 场景，合成一段平衡的 76 自由度全身加五指投掷轨迹，在闭环物理中验证 6/6 个目标全部被击倒，并导出 Blender 4.5 Cycles 渲染。

抱歉，您的浏览器不支持播放此视频

从草图到仿真——手绘草图到灵巧人形轨迹与 Cycles 渲染。

提示词：

[Attached: hand_drawn_throw_sketch.jpg]

Translate this hand-drawn sketch into a physically verified humanoid simulation in `mujoco_warp` and a raytraced Blender Cycles video:

1. Scene & Kinematic Setup from Sketch:
   - Parse the spatial layout from `hand_drawn_throw_sketch.jpg`: place the 76-DoF humanoid (32-DoF body equipped with two 22-DoF five-finger dexterous hands) at the origin facing a table holding a 3-2-1 pyramid of 6 rigid target boxes, with a projectile sphere grasped in the right five-finger hand.

2. Whole-Body Balance & Dexterous Throw Synthesis:
   - Synthesize a 50 Hz whole-body throwing trajectory (140 steps, 2.80 s) across four phases: (1) stable stance & 5-finger power grasp closure, (2) torso/shoulder wind-up with ankle-hip Center-of-Mass compensation, (3) forward whipping acceleration, and (4) synchronized 22-DoF five-finger opening at peak hand velocity so the ball launches on a ballistic arc into the pyramid's second-tier center of mass.
   - Verify in `mujoco_warp` contact physics that the humanoid maintains upright balance (< 3 cm pelvis drift) and knocks down all 6 target boxes.

纯文本

已复制

用 Argon 解锁物理理解

1. 事后标注与数据质量

Argon 对灵巧操作具有精确的物理理解，可以从视频中提供细致的分割和文本标注。Argon 能够识别较差的操作片段，例如停顿、不稳定的抓取和手忙脚乱的操作。给定一段双臂 ALOHA 2 机器人在 136 秒内将一张正方形纸折成折纸狐狸的俯拍视频，Argon 会生成一个时间对齐的 4 层事后标注层级：片段目标（G1）、6 个连续的语义子任务（G2）、17 个接触级原子基元（G3），以及自动质量控制（QC）重加权——在手忙脚乱和重试期间保持任务进度不下降。

抱歉，您的浏览器不支持播放此视频

Argon 多粒度事后标注与遥测 HUD。

提示词：

[Attached: Aloha2_Origami_20250309.mp4]

Produce a time-aligned multi-granularity hindsight annotation JSON with four levels:
1. G1_episode_instruction: Concise imperative sentence describing the overall manipulation goal.
2. G2_subtasks: Contiguous semantic subtask phases specifying which arm(s) anchor vs. fold/rotate the object.
3. G3_atomic_primitives: Fine-grained contact-level primitives (3–15s) aligned with grasp closure, crease sweeps, release, or retry attempts.
4. quality_control_and_reweighting: Hold task_progress_pct flat during non-progressing slip/delamination retry loops and downweight them (0.15x–0.25x) for Behavior Cloning.

纯文本

已复制

2. 找出异常的球

为了让机器人重现经典的 12 球天平谜题，Argon 构建了一个双臂 Franka FR3 的 MuJoCo 环境，其中包含 12 颗外观完全相同的台球（其中一颗暗地里更重或更轻）以及一个带关节的 Roberval 双盘天平。随后它编写了一个仅依赖视觉的闭环求解器：通过双目 RGB 相机对球进行三角定位，每次称重后通过视觉读取天平倾斜情况，并在 3 次称重内将异常球分离到重（红色）或轻（蓝色）托盘中。

抱歉，您的浏览器不支持播放此视频

找出异常球（纯视觉闭环求解器）：双臂 Franka FR3 在金色/青色秤盘上称量候选球，读取立体秤的倾斜，并在 3 次称量内将确认的偏重异常球隔离放入红色托盘。

提示词：

# Prompt 1 — Physics-Accurate MuJoCo Environment & Benchmark Generation
Make a MuJoCo environment where a dual-arm robot (two 7-DoF Franka FR3 arms with parallel-jaw grippers on a shared tabletop) is tasked to solve the classic 12-ball balance scale weight puzzle from https://www.mathsisfun.com/pool_balls_solution.html.
1. Create 12 visually indistinguishable pool balls in a 2x6 starting rack (11 standard 150 g, 1 odd ball either 270 g or 30 g).
2. Build a functional two-pan Roberval balance scale at the center of the table with parallelogram equality constraints (`mjEQ_JOINT`) so the Left (Gold) and Right (Teal) pans stay horizontal, plus Red (Heavy) and Blue (Light) submission trays.
3. Track distinct settled weighings (`weighing_count`) and reward `1.00` for isolating the true odd ball in the correct Heavy/Light tray within <= 3 weighings.

# Prompt 2 — Closed-Loop Vision-Only Solver
Implement an independent closed-loop policy relying strictly on stereo RGB camera inputs (`head_camera_left_rgb_image`, `head_camera_right_rgb_image`) and robot proprioception (`add_privileged_task_observations=False`):
1. Segment the 12 balls, Gold/Teal pans, and Red/Blue trays in both stereo views and triangulate their 3D coordinates.
2. Visually read the balance scale tilt after each weighing from stereo vertical pixel disparity (+1 = left heavier, -1 = right heavier, 0 = balanced).
3. Rotate the gripper wrist yaw by 90 degrees when grasping/placing inside the narrow scale pans to clear the raised rims, and isolate the odd ball into the Red or Blue tray within <= 3 weighings.

纯文本

已复制

3. 价值函数

Argon 是面向机器人任务的通用价值函数。在一个来自人类 UMI 数据集的 60 秒双臂整理环的片段上进行评估，Argon 围绕接触切换点迭代检查同步的头部、左夹爪和右夹爪相机画面，生成密集的价值函数 / 任务进度预测，并将轨迹分解为 28 个具有正、中、负进度斜率的片段。

抱歉，您的浏览器不支持播放此视频

价值函数：同步的三相机 UMI 回放（头部、左夹爪、右夹爪），与动画化的任务进度曲线及双臂片段标注锁定同步。

提示词：

[Attached: umi_head.mp4, umi_left_gripper.mp4, umi_right_gripper.mp4]

Look at 3 videos of human wearing a gripper performing activities from head, left, right gripper views. Look at the videos thoroughly and give a detailed plot of progress at the task (with positive, neutral, negative slope), identifying important segments in the video and labeling the activities. Look at individual frames or subsegments of the video if necessary to get a better understanding.

纯文本

已复制

用 Argon 解锁研究与自我改进

1. 通过仿真改进动作

Argon 可以在仿真中生成新数据，以提升通用视觉-语言-动作（VLA）模型的性能。从真实桌面的一张 RGB 照片出发，Argon 自主构建标定好的 MuJoCo 仿真孪生，定义新任务，并合成一个闭环运动生成器，在仿真中收集 4,500 多个域随机化片段——将真实世界物理机器人的成功率从 56% 提升到 75%，并将完成时间从 35 秒缩短到 24 秒。

抱歉，您的浏览器不支持播放此视频

通过仿真改进动作：提升了真实世界评估基准（成功率 56% → 75%，完成时间 35s → 24s）。

提示词：

[Attached: real_tabletop_photo.jpg]

You are given a single overhead RGB camera photo (`real_tabletop_photo.jpg`) of a bimanual robot workspace (two 7-DoF Franka FR3 arms with parallel-jaw grippers and 4 RGB cameras) facing a table arranged with photobooth party props.
1. Reconstruct procedural 3D meshes and compound MuJoCo collision primitives for all 9 party props matching their real-world dimensions, colors, and resting poses, including spring-hinged deformable crown panels for the soft plush cap.
2. Define a 77-task single-arm and bimanual curriculum (pick-and-place, uprighting, directional pushing, mid-air handovers, bimanual scoop-and-push, and pipelined table clearing).
3. Synthesize a closed-loop IK trajectory generator across ±10 cm pose randomization, verify `reward == 1.0` in MuJoCo physics, and export 4,500+ episodes for VLA fine-tuning.

纯文本

已复制

2. 视觉抓放任务中的自我改进

在一台简单的 6 自由度机械臂上，Argon 通过试错与自我反思，在抓放任务上迭代地同时改进执行速度和放置精度。

抱歉，您的浏览器不支持播放此视频

SO-101 视觉引导抓放——第 1 → 2 → 3 次迭代中的硬件自我改进：第 1 次迭代（58.9 秒）、第 2 次迭代（22.9 秒，投放时擦到箱沿）、第 3 次迭代（17.4 秒，正中投放）。

提示词：

# Stage 1 — Measure Objects & Dry-Run Plan (No Hardware Motion)
Stage 1: measure objects — find the objects and plan the whole motion. The arm does not move in this stage.

A. I put a LEGO brick and a box on the table in front of the arm, and the wooden box to one side. Find the brick and the wooden box in the image. And also in the real world. Show this in an annotated image.

B. A dry-run plan (no hardware). Write the planner to pick and place the lego block. Print the plan table, draw all poses onto the overhead image, and stop. Do not move the robot.
Report with evidence.

# Stage 2 — Pre-Pick Hover & Wrist-Camera Verification
Stage 2: Put the gripper above the brick without touching it, and verify with the wrist camera that the arm and the vision agree. Please record executions with both cameras (compressed video is OK).

# Stage 3 (Autonomous Pick-and-Place → Self-Improvement → Optimization):
1. Execute autonomously: Turn the Stage 1 dry run into a real grasp, lift, and place into the wooden box. Run it yourself subject to safety rules from AGENTS.md.

2. Self-improve: When an attempt fails, diagnose it yourself. Review your own logs and wrist/overhead footage, find the root cause, check the fix in a dry run, and retry.

3. Optimize on feedback: Once it succeeds, make it at least 2× faster within safety limits (rule R5) unless I authorize more, or give feedback.

纯文本

已复制

3. 在 MuJoCo WebAssembly 中用单条提示词复现机器人学论文

每篇论文只需一条提示词，Argon 即可阅读机器人控制论文及其发布的任意检查点，将物理模型和闭环控制器（从 DeepLoco 的 7 网络分层运动控制、HUSKY 的 50 Hz PPO 演员到 Whole-Body MPC 的 Web Worker iLQR）移植到自包含的 MuJoCo WebAssembly 浏览器实验环境中，并在无头 Chromium 中进行验证。

抱歉，您的浏览器不支持播放此视频

五篇里程碑式的足式与人形机器人论文在浏览器中实时运行（MuJoCo WASM）：Robotic Soccer OP3 1v1、DeepLoco、Massively Parallel Deep RL ANYmal C、Whole-Body MPC Go2 以及 HUSKY G1 滑板（出自 Argon 复现的 68 个交互式浏览器沙盒）。

提示词：

[Attached: paper.pdf, optional author_checkpoints/ or urdf_assets/]

Reproduce the paper "<Paper Title & arXiv ID>" from scratch in MuJoCo WebAssembly and deploy a self-contained browser lab + visual essay:
1. Match the paper's exact physics timestep, actuator gains (kp/kv), torque limits, and kinematic constraints in MuJoCo XML, and run real neural inference or Web Worker iLQR in JavaScript at the paper's exact control rates.
2. Build `index.html` (Spatial UI simulator with camera presets, live telemetry plots, and disturbance injection) and `scrollytelling.html` (interactive visual essay).
3. Run a headless Chromium (Playwright) verification loop asserting 0 console errors and non-zero task progress over 300+ frames.

纯文本

已复制

用 Argon 推进具身智能

Argon 融合了编程、视觉感知、物理仿真和具身机器人控制。通过对多视角视频、3D 几何、接触力学和本体感觉遥测数据进行原生推理，Argon 不仅能理解场景中有哪些物体，还能理解它们如何与智能体进行物理交互，以完成全身灵巧任务。Gemini 4 Argon 标志着加速迈向物理 AGI 的重要里程碑。通过构建这一核心智能，我们的目标是让机器人能够与人类并肩工作，解决复杂挑战。

开始使用

你今天就可以在 Google 的开发者平台上使用 Gemini 4 Argon 进行构建：

• Google AI Studio：面向 Google AI Ultra 订阅者开放。
• Gemini API：可在 AI Studio 中以 Gemini API 付费层级使用。在 Interactions API 中使用 gemini-4-argon，各 google-genai SDK 自首日起即提供支持。
• Google Antigravity：面向 Google AI Ultra 订阅者开放。要开始构建，请在 Google Antigravity 中选择 Gemini 4 Argon（gemini-4-argon）作为你的活跃模型，用于复杂规划、/goal 运行和多智能体编排。

上一页

下一页