--- title: Google Releases Gemini 4 Argon for Robotics Control,... description: Google released Gemini 4 Argon, a general-purpose physical-reasoning model for robot control, simulation, and policy improvement. It directs a Boston Dynamics... canonical: https://daily.dev/posts/google-releases-gemini-4-argon-for-robotics-control-simulation-and-policy-improvement-yzkewy6rg twitter:card: summary_large_image twitter:site: @dailydotdev og:type: website og:site_name: daily.dev og:title: Google Releases Gemini 4 Argon for Robotics Control, Simulation, and Policy Improvement | daily.dev og:description: Google released Gemini 4 Argon, a general-purpose physical-reasoning model for robot control, simulation, and policy improvement. It directs a Boston Dynamics... og:url: https://daily.dev/posts/google-releases-gemini-4-argon-for-robotics-control-simulation-and-policy-improvement-yzkewy6rg og:image: https://api.daily.dev/og/posts/yZkewy6rG.png og:image:alt: Google Releases Gemini 4 Argon for Robotics Control, Simulation, and Policy Improvement og:image:width: 1200 og:image:height: 630 og:locale: en --- > ## Documentation Index > Fetch the complete documentation index at: https://daily.dev/llms.txt > Use this file to discover all available pages before exploring further. # Google Releases Gemini 4 Argon for Robotics Control, Simulation, and Policy Improvement **[InfoQ](https://daily.dev/sources/infoq)** · 3 min read · 0 upvotes · 0 comments ## Summary Google released Gemini 4 Argon, a general-purpose physical-reasoning model for robot control, simulation, and policy improvement. It directs a Boston Dynamics Spot through a foot-pedal task after practicing in a MuJoCo digital twin, and steers a simpler arm using visual coordinates and HTTP motion primitives. Google reports a success rate rising from 56% to 75% on a real setup after fine-tuning a vision-language-action policy with synthesized domain-randomized MuJoCo episodes from a single RGB photo. Argon scores 87% on the ERQA benchmark using pure reasoning, and Google says that result leads other frontier models on that test. It's available to Google AI Ultra subscribers and through the paid Gemini API as gemini-4-argon. ## Full article daily.dev links to this article rather than hosting it. Read it at the original source: ## Questions this post answers ### What is Gemini 4 Argon used for in robotics? Gemini 4 Argon is Google's general-purpose physical-reasoning model for robot control, Real2Sim scene construction, multimodal labeling of robot video, and iterative policy improvement, rather than a single end-effector policy. It has directed a Boston Dynamics Spot through a foot-pedal task after practicing in a MuJoCo digital twin, and steered a simpler arm using visual coordinates and HTTP motion primitives. _Teams evaluating physical-reasoning models for robotics can track Gemini 4 Argon coverage and benchmarks on daily.dev._ ### How well does Gemini 4 Argon perform on the ERQA robotics benchmark? Gemini 4 Argon scores 87% on the Embodied Reasoning Question Answering (ERQA) benchmark using pure reasoning without helper tools, a result Google says leads other frontier models on that test. The model is available to Google AI Ultra subscribers and through the paid Gemini API under the model id gemini-4-argon. _Developers comparing frontier robotics models can follow ERQA benchmark results via daily.dev._ ### Can synthetic MuJoCo simulation data improve real robot success rates? Yes, Google reports synthesizing domain-randomized MuJoCo episodes from a single RGB photo to fine-tune a vision-language-action policy, which raised success rate from 56% to 75% on the corresponding real setup. This demonstrates Real2Sim data generation can meaningfully boost real-world robot policy performance without extensive teleoperation data collection. _Those building sim-to-real pipelines can keep up with Real2Sim technique results on daily.dev._ ## Similar posts on daily.dev - [Gemini 4 Argon: Google’s New AI Is Built to Think and Work for Much Longer](https://daily.dev/posts/gemini-4-argon-google-s-new-ai-is-built-to-think-and-work-for-much-longer-ibbgfhpdj) · Medium · 0 upvotes · 0 comments --- Tags: [#google](https://daily.dev/tags/google), [#robotics](https://daily.dev/tags/robotics), [#google-gemini](https://daily.dev/tags/google-gemini) [View this post on daily.dev](https://daily.dev/posts/google-releases-gemini-4-argon-for-robotics-control-simulation-and-policy-improvement-yzkewy6rg) ```json {"@context":"https://schema.org","@graph":[{"@type":"Organization","@id":"https://daily.dev/#organization","name":"daily.dev","url":"https://daily.dev","logo":{"@type":"ImageObject","url":"https://daily.dev/apple-touch-icon.png","width":180,"height":180},"sameAs":["https://twitter.com/dailydotdev","https://github.com/dailydotdev","https://www.linkedin.com/company/daily-dev-ltd"]},{"@type":"WebSite","@id":"https://daily.dev/#website","url":"https://daily.dev","name":"daily.dev","publisher":{"@id":"https://daily.dev/#organization"},"potentialAction":{"@type":"SearchAction","target":{"@type":"EntryPoint","urlTemplate":"https://daily.dev/search?q={search_term_string}"},"query-input":"required name=search_term_string"}}]} {"@context":"https://schema.org","@type":"TechArticle","headline":"Google Releases Gemini 4 Argon for Robotics Control, Simulation, and Policy Improvement","url":"https://daily.dev/posts/google-releases-gemini-4-argon-for-robotics-control-simulation-and-policy-improvement-yzkewy6rg","mainEntityOfPage":{"@type":"WebPage","@id":"https://daily.dev/posts/google-releases-gemini-4-argon-for-robotics-control-simulation-and-policy-improvement-yzkewy6rg"},"datePublished":"2026-11-09T15:47:52.398Z","dateModified":"2026-11-10T16:09:03.065Z","description":"Google released Gemini 4 Argon, a general-purpose physical-reasoning model for robot control, simulation, and policy improvement. It directs a Boston Dynamics...","image":"https://media.daily.dev/image/upload/f_auto,q_auto/v1/posts/4ffd6c4951a90bee2d01d6df1f98e535?_a=AQAEuop","thumbnailUrl":"https://media.daily.dev/image/upload/f_auto,q_auto/v1/posts/4ffd6c4951a90bee2d01d6df1f98e535?_a=AQAEuop","isAccessibleForFree":true,"articleSection":"InfoQ","inLanguage":"en","publisher":{"@type":"Organization","name":"daily.dev","url":"https://daily.dev","logo":{"@type":"ImageObject","url":"https://daily.dev/apple-touch-icon.png","width":180,"height":180}},"author":{"@type":"Organization","name":"InfoQ","logo":"https://media.daily.dev/image/upload/t_logo,f_auto/v1/logos/afc3bced3e1e4b188dd9127017a60e0c","url":"https://daily.dev/sources/infoq"},"commentCount":0,"discussionUrl":"https://daily.dev/posts/google-releases-gemini-4-argon-for-robotics-control-simulation-and-policy-improvement-yzkewy6rg","interactionStatistic":[{"@type":"InteractionCounter","interactionType":{"@type":"LikeAction"},"userInteractionCount":0},{"@type":"InteractionCounter","interactionType":{"@type":"CommentAction"},"userInteractionCount":0}],"keywords":"google,robotics,google-gemini","timeRequired":"PT3M"} {"@context":"https://schema.org","@type":"BreadcrumbList","itemListElement":[{"@type":"ListItem","position":1,"name":"Home","item":"https://daily.dev"},{"@type":"ListItem","position":2,"name":"InfoQ","item":"https://daily.dev/sources/infoq"},{"@type":"ListItem","position":3,"name":"Google Releases Gemini 4 Argon for Robotics Control, Simulation, and Policy Improvement"}]} {"@context":"https://schema.org","@type":"FAQPage","@id":"https://daily.dev/posts/google-releases-gemini-4-argon-for-robotics-control-simulation-and-policy-improvement-yzkewy6rg#faq","mainEntity":[{"@type":"Question","name":"What is Gemini 4 Argon used for in robotics?","acceptedAnswer":{"@type":"Answer","text":"Gemini 4 Argon is Google's general-purpose physical-reasoning model for robot control, Real2Sim scene construction, multimodal labeling of robot video, and iterative policy improvement, rather than a single end-effector policy. It has directed a Boston Dynamics Spot through a foot-pedal task after practicing in a MuJoCo digital twin, and steered a simpler arm using visual coordinates and HTTP motion primitives. Teams evaluating physical-reasoning models for robotics can track Gemini 4 Argon coverage and benchmarks on daily.dev."}},{"@type":"Question","name":"How well does Gemini 4 Argon perform on the ERQA robotics benchmark?","acceptedAnswer":{"@type":"Answer","text":"Gemini 4 Argon scores 87% on the Embodied Reasoning Question Answering (ERQA) benchmark using pure reasoning without helper tools, a result Google says leads other frontier models on that test. The model is available to Google AI Ultra subscribers and through the paid Gemini API under the model id gemini-4-argon. Developers comparing frontier robotics models can follow ERQA benchmark results via daily.dev."}},{"@type":"Question","name":"Can synthetic MuJoCo simulation data improve real robot success rates?","acceptedAnswer":{"@type":"Answer","text":"Yes, Google reports synthesizing domain-randomized MuJoCo episodes from a single RGB photo to fine-tune a vision-language-action policy, which raised success rate from 56% to 75% on the corresponding real setup. This demonstrates Real2Sim data generation can meaningfully boost real-world robot policy performance without extensive teleoperation data collection. Those building sim-to-real pipelines can keep up with Real2Sim technique results on daily.dev."}}]} ```