{"slug": "mujoco", "name": "MuJoCo", "category": "other", "category_name": "Other software", "description": "", "recommenders": 3, "commenters": 8, "warners": 1, "score": 2.954, "rising": -0.98, "first": "2026-03-05", "latest": "2026-10-27", "why": "The best thing is that we do not need to fight with NVIDIA software and that it runs on macOS.", "why_is_quote": true, "has_verdict": false, "sentiment": {"recommend": 3, "neutral": 4, "compare": 0, "warn": 1}, "spark": [0, 0, 0, 0, 0, 2, 1, 0, 0, 0, 0, 0], "listed": true, "loved": [], "caveat": "", "short_quote": "The best thing is that we do not need to fight with NVIDIA software and that it runs on macOS.", "kind": "", "icon": true, "domain": null, "github_repo": "github.com/google-deepmind/mujoco", "official_url": "https://mujoco.org", "pricing": "", "aliases": ["MuJoCo"], "verdict": null, "praise": [{"label": "", "summary": "", "quotes": [{"id": 771211216, "item_id": 48200701, "url": "https://news.ycombinator.com/item?id=48200701", "by": "jackling", "date": "2026-05-19", "version": null, "stance": "recommend", "quote": "MuJoCo is the goto here", "excerpt": "Would be similar to the typical simulations of humanoids. If you need to model the deformations of the human body, or get a proper model of tendons that make up humans, it'll be more difficult, but possible. Proper simulators for those exist, you essentially need an engine with a compliant contact model. MuJoCo is the goto here, see: https://mujoco.readthedocs.io/en/stable/modeling.html#muscles https://mujoco.readthe", "context": "works in rigid body simulation for robotics", "use_case": "Physics simulation engine", "switched_from": "", "story": "Gemini Omni"}, {"id": 765800208, "item_id": 47862513, "url": "https://news.ycombinator.com/item?id=47862513", "by": "tlb", "date": "2026-04-22", "version": null, "stance": "recommend", "quote": "MuJoCo is great.", "excerpt": "MuJoCo is great. I have it running in the browser for robotics simulation. See for example https://visibot.com/sheet/examples/humanoid_walking.v", "context": "", "use_case": "Robotics simulation", "switched_from": "", "story": "MuJoCo \u2013 Advanced Physics Simulation"}, {"id": 765787472, "item_id": 47861717, "url": "https://news.ycombinator.com/item?id=47861717", "by": "prathje", "date": "2026-04-22", "version": null, "stance": "recommend", "quote": "The best thing is that we do not need to fight with NVIDIA software and that it runs on macOS.", "excerpt": "We are using MuJoCo to train a G1 humanoid robot right now. The best thing is that we do not need to fight with NVIDIA software and that it runs on macOS. PS: I just finished a first draft for agentic skills around working with MuJoCo in Python. Feel free to check them out here: https://github.com/prathje/agentic_mujoco_skills", "context": "", "use_case": "robot training simulation", "switched_from": "", "story": "MuJoCo \u2013 Advanced Physics Simulation"}]}], "caveats": [{"label": "", "summary": "", "quotes": [{"id": 756281826, "item_id": 47267614, "url": "https://news.ycombinator.com/item?id=47267614", "by": "ZuoCen_Liu", "date": "2026-03-05", "version": null, "stance": "warn", "quote": "simulations from engines like Isaac Sim, MuJoCo, or world models like Marble often fail the moment they are deployed on real robotic hardware", "excerpt": "I've been obsessed with a single question: Why do \"visually stunning\" simulations from engines like Isaac Sim, MuJoCo, or world models like Marble often fail the moment they are deployed on real robotic hardware? The Theory: I am proposing the Non-Associative Residual Hypothesis (NARH). By auditing 7-DoF trajectories, we can detect the underlying physical consistency\u2014or lack thereof\u2014in these systems. In plain terms: ", "context": "", "use_case": "", "switched_from": "", "story": "SIPA: Auditing Physical Integrity in World Models and Robotics(Isaac SIM,Marble)"}]}], "who": [{"by": "jackling", "context": "works in rigid body simulation for robotics", "url": "https://news.ycombinator.com/item?id=48200701", "date": "2026-05-19"}], "tags": [], "use_cases": ["robotics simulation", "robot simulation for rl", "robot simulation rendering", "physics simulation engine", "robot policy training sim", "robot training simulation"], "compared": [], "makers": [{"id": 765837216, "item_id": 47864826, "url": "https://news.ycombinator.com/item?id=47864826", "by": "saran-t", "date": "2026-04-22", "version": null, "stance": "recommend", "quote": "we're also working on an official browser-based interactive viewer", "excerpt": "While we have people's attention here, we're also working on an official browser-based interactive viewer that allows people to specify a URL to their own model and share a link. Very early experimental prototype at the moment, doesn't work super well on phones yet, but you can try it at https://saran-t.github.io/mujoco/?model=https://raw.githubusercontent.com/google-deepmind/mujoco/main/model/cards/cards.xml It'll b", "context": "", "use_case": "", "switched_from": "", "story": "MuJoCo \u2013 Advanced Physics Simulation"}], "threads": [{"url": "https://news.ycombinator.com/item?id=47820791", "title": "MuJoCo \u2013 Advanced Physics Simulation", "mentions": 4}, {"url": "https://news.ycombinator.com/item?id=50258193", "title": "CarveAnything: Image-to-volumetric material with destruction mechanics", "mentions": 1}, {"url": "https://news.ycombinator.com/item?id=49462763", "title": "Microduck", "mentions": 1}, {"url": "https://news.ycombinator.com/item?id=49449201", "title": "Show HN: A robot football league where frontier AI models manage the clubs", "mentions": 1}, {"url": "https://news.ycombinator.com/item?id=48196609", "title": "Gemini Omni", "mentions": 1}, {"url": "https://news.ycombinator.com/item?id=47267570", "title": "SIPA: Auditing Physical Integrity in World Models and Robotics(Isaac SIM,Marble)", "mentions": 1}], "sentiment_by_month": {"2025-11": {}, "2025-12": {}, "2026-01": {}, "2026-02": {}, "2026-03": {"warn": 1}, "2026-04": {"recommend": 2, "neutral": 1}, "2026-05": {"recommend": 1}, "2026-06": {}, "2026-07": {}, "2026-08": {"neutral": 2}, "2026-09": {}, "2026-10": {"neutral": 1}}, "versions": [], "version_unnamed": 0, "family_ref": null, "runs_on": null, "lines": [], "quotes": [{"id": 804131106, "item_id": 50258194, "url": "https://news.ycombinator.com/item?id=50258194", "by": "lukehollis", "date": "2026-10-27", "version": null, "stance": "neutral", "quote": "simulation environments in Isaac Lab or MuJoCo", "excerpt": "CarveAnything takes an input monocular image and generates a 3d model with predictive volumetric content, physics properties, and destruction mechanics for simulating cutting, lighting, and staging objects--so it may be adapted to the Halloween season. One can use this to automate robotics policy evaluation and training on diverse objects, and more on methods is in the paper. We use a combination of semantic features", "context": "", "use_case": "robotics simulation", "switched_from": "", "story": "CarveAnything: Image-to-volumetric material with destruction mechanics"}, {"id": 791464256, "item_id": 49466516, "url": "https://news.ycombinator.com/item?id=49466516", "by": "prometheus1992", "date": "2026-08-27", "version": null, "stance": "neutral", "quote": "most of the robotics stuff that you see in the news is based on MujoCo", "excerpt": "Not sure many people know this - most of the robotics stuff that you see in the news is based on MujoCo (Multi-Joint dynamics with Contact) engine, that is used primarily to create simulated environments in which these robots learn the RL policies. This project is maintained by Google deepmind - https://github.com/google-deepmind/mujoco", "context": "", "use_case": "robot simulation for RL", "switched_from": "", "story": "Microduck"}, {"id": 791188593, "item_id": 49449287, "url": "https://news.ycombinator.com/item?id=49449287", "by": "spobin", "date": "2026-08-26", "version": null, "stance": "neutral", "quote": "decided to play around with the MuJoCo rendering system", "excerpt": "I've been fascinated by Unitree's G1 robot architecture and, without the $20k to buy a unit, decided to play around with the MuJoCo rendering system. I fell down a bit of a rabbit hole and ended up with an automated football league that runs 3 matches per day and streams on Twitch/Youtube. It's been a lot of fun! Four of the eight clubs are managed by frontier AI models (Claude, GPT, Gemini, Manus). Between match day", "context": "", "use_case": "robot simulation rendering", "switched_from": "", "story": "Show HN: A robot football league where frontier AI models manage the clubs"}, {"id": 771211216, "item_id": 48200701, "url": "https://news.ycombinator.com/item?id=48200701", "by": "jackling", "date": "2026-05-19", "version": null, "stance": "recommend", "quote": "MuJoCo is the goto here", "excerpt": "Would be similar to the typical simulations of humanoids. If you need to model the deformations of the human body, or get a proper model of tendons that make up humans, it'll be more difficult, but possible. Proper simulators for those exist, you essentially need an engine with a compliant contact model. MuJoCo is the goto here, see: https://mujoco.readthedocs.io/en/stable/modeling.html#muscles https://mujoco.readthe", "context": "works in rigid body simulation for robotics", "use_case": "Physics simulation engine", "switched_from": "", "story": "Gemini Omni"}, {"id": 765850560, "item_id": 47865660, "url": "https://news.ycombinator.com/item?id=47865660", "by": "saucesaft", "date": "2026-04-22", "version": null, "stance": "neutral", "quote": "I actually worked with getting MuJoCo's physics gradients to train quadruped locomotion policies", "excerpt": "My two cents, I actually worked with getting MuJoCo's physics gradients to train quadruped locomotion policies. Really interesting field. https://github.com/saucesaft/differential_policies/", "context": "", "use_case": "Robot policy training sim", "switched_from": "", "story": "MuJoCo \u2013 Advanced Physics Simulation"}, {"id": 765800208, "item_id": 47862513, "url": "https://news.ycombinator.com/item?id=47862513", "by": "tlb", "date": "2026-04-22", "version": null, "stance": "recommend", "quote": "MuJoCo is great.", "excerpt": "MuJoCo is great. I have it running in the browser for robotics simulation. See for example https://visibot.com/sheet/examples/humanoid_walking.v", "context": "", "use_case": "Robotics simulation", "switched_from": "", "story": "MuJoCo \u2013 Advanced Physics Simulation"}, {"id": 765787472, "item_id": 47861717, "url": "https://news.ycombinator.com/item?id=47861717", "by": "prathje", "date": "2026-04-22", "version": null, "stance": "recommend", "quote": "The best thing is that we do not need to fight with NVIDIA software and that it runs on macOS.", "excerpt": "We are using MuJoCo to train a G1 humanoid robot right now. The best thing is that we do not need to fight with NVIDIA software and that it runs on macOS. PS: I just finished a first draft for agentic skills around working with MuJoCo in Python. Feel free to check them out here: https://github.com/prathje/agentic_mujoco_skills", "context": "", "use_case": "robot training simulation", "switched_from": "", "story": "MuJoCo \u2013 Advanced Physics Simulation"}, {"id": 756281826, "item_id": 47267614, "url": "https://news.ycombinator.com/item?id=47267614", "by": "ZuoCen_Liu", "date": "2026-03-05", "version": null, "stance": "warn", "quote": "simulations from engines like Isaac Sim, MuJoCo, or world models like Marble often fail the moment they are deployed on real robotic hardware", "excerpt": "I've been obsessed with a single question: Why do \"visually stunning\" simulations from engines like Isaac Sim, MuJoCo, or world models like Marble often fail the moment they are deployed on real robotic hardware? The Theory: I am proposing the Non-Associative Residual Hypothesis (NARH). By auditing 7-DoF trajectories, we can detect the underlying physical consistency\u2014or lack thereof\u2014in these systems. In plain terms: ", "context": "", "use_case": "", "switched_from": "", "story": "SIPA: Auditing Physical Integrity in World Models and Robotics(Isaac SIM,Marble)"}]}