OCT. 13, 2026 (英语).
Gemini 4 Argon标志着物理理解、代码生成和视觉语义推理的一步变化——将新的工作流程解锁在机器人堆栈上，并加速向物理AGI迈进。
您可以在健美理性(ERQA)基准测试上看到这种能力的飞跃，测试一个模型在3D环境中理解和解决物理，空间问题的能力。 使用纯推理，阿尔贡得分87%，优于所有其他前沿模型。
作为通用物理推理模型，Argon可以被应用到整个机器人生命周期中。 在此文中，我们分享了阿尔贡在四个关键领域的代表意向：
- 机器人控制移动操纵与MuJoCo练习，零镜头机器人控制，以及视频世界模型的闭路引导。
——Real2Sim 从机器人相机种子中重建物理校准的模拟双胞胎，以自我为中心的人类视频，或手绘纸草图。
- 物理理解
- 通过模拟和硬件迭代自我改进来研究和自我改进行动。
Argon控制一个波士顿动态(Boston Dynamics Spot®)四重奏(12-DoF腿+6-DoF手臂)来学习新技能：踩踏脚踏板垃圾桶丢弃汽水罐。 Argon在踩踏物理踏板和将汽水罐扔进打开的垃圾桶之前，会创建一个清晰的MuJoCo数字双胞胎脚踏板垃圾桶来练习其动作。
斑点移动操纵——Real-to-Sim-to-Real脚踏车垃圾桶处理：斑点捡起汽水罐，检查其手腕相机，在MuJoCo练习三脚架平衡，踩脚踏板，验证手放在开箱处，把罐子放进去。
提示：
左边的黑桌子上还有一个
使用即时移动操纵技术来捡起来，然后把它们丢进黑色的脚踏板箱。
在踏板上踩踏前，在前摄像头中验证你的前右脚目标以踏板为中心，并使用2-agent验证集群(脚检和手检)在释放前同时验证脚踏板步骤和3D手套对齐。
纯文本
复制
Argon控制着一个简单的机器人手臂将一个掉下来的杯子直立地放进去，然后把餐巾放进去。 使用视觉坐标带和动作原始的Argon理由，在几个转弯执行这项任务。
左侧：第三人称真机器人视图；右侧：校准手腕相机，投射2厘米世界坐标和下巴闭合回旋器。
[附：手腕摄像头.png,obs.json].
任务：让机器人手臂把杯子右转，把餐巾放在杯子里。
□ 执行纪律
- 简单的产出1句，紧接着是你们的`curl ' 命令(第1轮和QQ想法：`每一轮)。 所有需要都用`obs.json ' 和`wrist cam.png ' 上绘制的校准电阻/电网。
□视觉坐标参考(手腕上注解 cam.png).
远离机器人基地 `-(DOWN)：走向机器人基地
世界左转右转
- `Z '：表上的高度`黄色数字 '：世界( X, Y) 厘米( 2 cm 网格 at `grid z')
- " 青色+横纹 "：目前的3D剪接中心
- `青色[+]括号'：当前下巴开口和关闭线
- " 绿线 "：当前物体长轴方向(`yaw deg ':0° =垂直沿，90° =水平沿+Y)
- `pitch deg=90°`：直下(-Z)自上而下抓住/位置；`pitch deg < 60°`：沿指轴倾斜的同轴方法
□ 闭环旋转工作流程
1. 检查当前状态和视图`wrist cam.png':
`curl - s http://127.0.0.1:8081/obs ' (返回`ee pos '、`持有 '、`地位 '、`工作空间 '、`wrist cam path')。
2. 选择 VLM 调用前方的颗粒度：
- 粗体(2至4个电话):`看'-`grasp '-`place '-`done '
- Grasp前核查(4-7次呼叫):`看 ' - ' goto(z=0.04) ' (核实氰 ' [+] ' 括号横跨目标)- ' grasp ' - > ' 位置直到接触 ' - > `done '
- 精细视觉攻击(8至20次通话):`看 ' - ' ' `grasp ' - ' 前置`goto ' - ' ' 闭路 ' by ' /`move until contact ' - ' done '
3. 发送运动原始(`POST/Plan ' )，并在每次转弯后在更新的`wrist cam.png ' 中进行视觉验证。 如果`状态================
□ 守则-即时政策先导(`POST http://127.0.0.1:8081/plan')
机器人。 look at(x=0.35,y=0.13,z=0.14，网格 z=0.015)
x,y=机器人.deproject(u=0.52),v=0.48,z 飞机=0.015
机器人.grasp(x=x,y=y,z=0.015,yaw deg=-35,pitch deg=90)
机器人.place(x=0.24,y=0.06,z=0.075)
机器人.place 直到接触(x=0.24,y=0.06,z=0.015,max lag=0.012,relief=0.02)
robot.goto(x=0.28,y=0.05,z=0.04,arc z=0.14, pitch deg=25,azimouth deg=90,roll deg=0，速度=0.10).
robot.move by(向前=0.025,dz=-0.02,dpitch deg=-3,droll deg=5，速度=0.04) 互联网档案馆的存檔，存档日期2013-07-02.
robot.move 直到 contact(dz=-0.06)，速度=0.03，最大 lag=0.012, relief=0.02).
机器人.open gripper () / 机器人。 close gripper () / 机器人。 home () / 机器人.done ()
Argon在Gemini Omni World Model中引导机器人将五块分散的拼音瓦片(G,R,A,N,O)重新排列成干净的行拼音A-R-G-O-N. Argon不指定一个静态的快板，而是在闭合的视觉环中引导。 它写有时间标记的行动提示，检查中间推出的键框，并完善它的提示。
迭代视频世界模式指导 Dexterous Hands:33-Step指导视频推出拼写 A-R-G-O-N.
目标：{任务目标}
您正在引导一个视频世界模型( Gemini Omni) 在迭代视觉反馈循环中。 查看最近推出的附图框，并为视频世界模型写下一次加印动作提示：
- 一次把复杂的运动打成一个简单的单向动作。
- 保持不活跃的手收回，避免挡住相机对工作空间内物体的视线。
- 明确说明物体移动后工作空间的哪些区域变为空。
- 在每次动作节拍时使用[MM:SS.mmm]时间戳，并在<omni prompt>标记中将输出包裹。
示例格式( 仅样本时间戳，而非全部提示):
<omni_prompt>
[00: 00.000] 字符A的左手停留在侧面，而右手伸向目标物体。
[00:02.500] 字符A的右指把物体顺利地推入表的中心，使其起始区域空空。
[00:05.000] 字符A的右手打开并收回到侧面，这样所有放置的物体都保持清晰的镜头视野。
</omni prompt> (英语).
Argon查看一个真正的机器人远程操作插曲，并发展出一个数字双胞胎；它获得或创建模拟资产，在与狡猾的手接触时看起来和行为都像真正的物体。 Argon调制资产和场景，以迭代视觉推理循环中真实录制数据中的模拟视觉和接触事件相匹配。
左(Real View)：同步真实的机器人相机种子；右(模拟):3D MuJoCo场景由Argon重建并校准。
[附图：机器人推出集
构建机器人的MuJoCo场景并重建任务对象。 当每集都按帧重播时，将机器人关节推进到记录的观察中。 状态与联合空间PD。
采用两个核查目标：
(1)视觉对接：MuJoCo从头左，左左，右右-右相机上排出这些真实的观察，并看起来像。
(2)物理匹配：在传感器框中，每个Sharpa指尖上的模拟接触力/齿轮在时间，尺寸和方向上匹配记录的触觉力。
您可以在模拟开始前调音：
(1)静态物体资产：现实几何，碰撞形状，质量，惯性，摩擦和关节，外观接近实际物体，加上初始物体在t=0.
(2)离线计划机器人指令，仅在真指尖靠近物体时在框中更改。 每个命令必须保持在已登录的遥波轨迹的0.55弧度(臂关节)和0.65弧度(手关节)以内。 其它任何地方，重放记录轨迹不变。
没有sim作弊：任务对象只能通过真正的MuJoCo接触机器人手和环境来移动(mj step).
t=0后不允许：更改MjModel参数；对对象施加外部力或牵引力；焊接或等式约束；在联合空间PD之外向导牵引力；设置或回滚对象qpos/qvel；致残或排除碰撞；报告不来自MuJoCo解析器的接触力(例如复制记录的触觉读数).
立即将重建的场景和推出结果作为交互HTML会话查看器，每集一页，显示：
- 在底部：一个同步的时间线(游戏、洗涤、速度控制)，驱动每个面板。
- 在左面板上：一个交互式的3D Blender场景，显示模拟机器人和物体，显示联系点和力箭头作为重播进度。
- 在右边的面板上：真实的 vs. sim 相机并肩对着所有 3 相机。 真实对准sim指力可视化。
鉴于一个以自我为中心的人类操纵任务的视频，Argon识别了与任务相关的物体，用复合凸轮脱落的夹层取代了用于凸轮容器的粗糙的捆绑箱，并将人类的手轨重新瞄准一个76-DoF双人体，在MuJoCo接触物理中用22-DoF五指手。
人景到模拟——Egocentre Human Video to Dexterous Humanoid Twin：左侧显示输入自我中心人类RGB视频；右侧显示同步500Hz MujoCo + Blender Cycles 多视图模拟推出。
[附图：以自我为中心的 human video.mp4，候选 sim views.png，手 3d keypoints.json].
您正在构建一个物理溶解的500 Hz MujoCo 操纵环境(`scene.xml')和工作室 Blender 的渲染，它忠实地复制了用22-DoF 解锁的五指手的双人形机器人上的`egocent human video.mp4'显示的桌面任务：
1. 场景重建和相机校准：
- 仅确定与操纵有关的物体(被抓住的目标物体和接收的容器或地标)，并将其3D平面坐标从自我中心照相机内在/外观中三角化。
- 将粗糙的捆绑箱替换成复合凸轮脱落的3D meshes，这样可以装入容器(被抹掉的碗、嵌入的木制托盘、开箱)，支持室内实际放置。
2. 500 Hz Dexterous Motion-Fiting压力测试：
- 将3D人的手腕和手指键点重新瞄准人造物的7-DoF臂和22-DoF手关节，在500赫兹(dt = 0.002s)以零动能的焊接约束下步入MuJoCo的轨迹。
- 核查对立的拇指和手指的关闭、干净的物体升力( > 2厘米)以及目标地标内部或附近稳定的最终位置。
与Argon一起，开发者可以拍下一张粗糙的笔画和纸画草图的照片——显示阿波罗2号人造物沿着抛物线弧向桌子上投掷一个球，将一个3-2-1堆的箱子顶上——并让Argon自动构造GPU加速的mujoco warp场景，合成一个平衡的76-DoF全体和五指抛射轨迹，验证6/6目标在闭路物理中击落，并导出一个Blender 4.5循环渲染。
缩写到模拟——手绘到Dexterous Humanoid Tropeory & Cycles Render的缩写。
[附图：手拖推 sk.
将这张手绘的草图翻译为“mujoco warp”中经过物理验证的人体模拟
1. Sketch的场景和动因设置：
- 分析空间布局，取自`Hand drawn throw sketch.jpg'：将76-DOF人形(32-DOF机体，配备两只22-DOF五指叉手)放置在正对着一个3-2-1金字塔6个刚性目标箱的桌子的原位上，右五指手握着一个射弹球。
2. 整体-Body
本文由 AI宝盒自动抓取并翻译，版权归原作者所有；如需查看原文请点下方链接。
OCT. 13, 2026
Gemini 4 Argon marks a step change in physical understanding, code generation, and visual-semantic reasoning—unlocking new workflows across the robotics stack and accelerating progress toward physical AGI.
You can see this leap in capability on the Embodied Reasoning (ERQA) benchmark, which tests a model's ability to understand and solve physical, spatial problems in 3D environments. Using pure reasoning, Argon scores 87%, outperforming every other frontier model.
As a general-purpose physical reasoning model, Argon can be applied across the entire robotics lifecycle. In this post, we share representative prompts for Argon across four key areas:
- Robot control → Mobile manipulation with MuJoCo practice, zero-shot robot control, and closed-loop steering of video world models.
- Real2Sim → Reconstructing physically calibrated simulation twins from robot camera feeds, egocentric human video, or hand-drawn paper sketches.
- Physical understanding → Agentic multimodal hindsight labeling including detailed understanding of dexterity, dense value functions and progress estimation.
- Research and self-improvement → Action improvement via simulation and iterative self-improvement on hardware.
Argon controls a Boston Dynamics Spot® quadruped (12-DoF legs + 6-DoF arm) to learn a new skill: stepping on a foot-pedal trash bin to throw away soda cans. Argon creates an articulated MuJoCo digital twin of a foot-pedal trash bin to practice its moves before stepping on the physical pedal and dropping soda cans into the open bin.
Spot Mobile Manipulation — Real-to-Sim-to-Real foot-pedal trash bin disposal: Spot picking up the soda can, checking its wrist camera, practicing 3-leg tripod balance in MuJoCo, stepping on the foot pedal, verifying hand centering in the open bin, and dropping the can inside.
Prompt:
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.
Plain text
Copied
Argon controls a simple robotic arm to place a fallen mug upright and put the crumpled napkin inside. Argon reasons with a visual coordinate harness and action primitives to perform this task over several turns.
Left: 3rd-person real robot view; Right: calibrated wrist camera with projected 2 cm world coordinates and jaw-closing reticle.
[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()
Argon steers robots in the Gemini Omni World Model to rearrange five scattered Scrabble tiles (G, R, A, N, O) into a clean row spelling A - R - G - O - N. Instead of specifying one static prompt, Argon steers in a closed visual loop. It writes timestamped action prompts, inspecting intermediate rollout keyframes, and refining its prompts.
Iterative Video World Model Steering Dexterous Hands: 33-Step Steered Video Rollout Spelling A-R-G-O-N.
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>
Argon views a real robot teleoperation episode and develops a digital twin; it obtains or creates simulation assets that look and behave like the real objects upon contact with the dexterous hands. Argon tunes the assets and scenes to match sim visuals and contact events with those in the real recorded data in an iterative visual-reasoning loop.
Left (Real View): synchronized real robot camera feeds; Right (Simulated): 3D MuJoCo scene reconstructed and calibrated by Argon.
[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.
Given an egocentric human video of a manipulation task, Argon identifies the task-relevant objects, replaces crude bounding boxes with compound convex-decomposed meshes for concave containers, and retargets the human hand trajectory onto a 76-DoF bimanual humanoid with 22-DoF five-fingered hands in MuJoCo contact physics.
Human View to Simulation — Egocentric Human Video to Dexterous Humanoid Twin: Left shows the input egocentric human RGB video; Right shows the synchronized 500 Hz MuJoCo + Blender Cycles multi-view simulation rollout.
[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.
With Argon, a developer can snap a photo of a rough pen-and-paper sketch—showing an Apollo 2 humanoid throwing a ball along a parabolic arc to topple a 3-2-1 stack of boxes on a table—and have Argon automatically construct the GPU-accelerated mujoco_warp scene, synthesize a balanced 76-DoF whole-body and five-finger throw trajectory, verify 6/6 target knockdown in closed-loop physics, and export a Blender 4.5 Cycles render.
Sketch to Simulation — Hand-Drawn Sketch to Dexterous Humanoid Trajectory & Cycles Render.
[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
Auto-fetched and translated by AI Gem Box. All rights belong to the original author.