Fourier Intelligence
Fourier GR-1
- Autonomy
- IIAssisted
- Pre-release
- 75/100Promising
- Readiness
- Not assessed
- Price
- Not tracked
ComparisonFourier GR-1 and Nori A3
Nori Robotics
Nori A3
- Autonomy
- IIAssisted
- Pre-release
- 51/100Plausible
- Readiness
- Not assessed
- Price
- $1,688list
Up to three
Autonomy Ladder
- IManual
- IIAssisted
- IIIConditional
- IVEnvironmental
- VGeneralized
All humanoid robots. Every figure below is placed against the whole class, so a gap can be read against how much the class itself varies.
Where each one suits you.
The conditions each robot was assessed against, side by side. Rows marked in gold are where these machines genuinely differ.
| Condition | Fourier GR-1 | Nori A3 |
|---|---|---|
| Where it operates | ||
| Home / residentialthey differ | PoorNot available for home use; designed for institutional and industrial deployment only. | DependsLaunch demos show home chores, but no delivered-unit report shows them working without the owner's own training. |
| Office / retail | DependsRequires factory-floor integration; not plug-and-play for general commercial environments. | DependsOffice and shelf-restocking use appears in Nori's Y Combinator profile with no published demonstration behind it. |
| Warehouse / factorythey differ | GoodDesigned for industrial logistics, inspection, and assembly with on-site technician support. | PoorEach arm is rated for 1.5 kg on a differential-drive base; nothing documents industrial duty. |
| Primary tasks | ||
| Household choresthey differ | PoorNo consumer household robotics support; dexterity is optimized for industrial manipulation, not domestic tasks. | DependsFetching, pouring and folding appear in launch footage; each task must be recorded in Nori Lab before it runs. |
| Social / carethey differ | PoorTeleoperation dependency and institutional-only deployment excludes social-companion use cases. | DependsA microphone and speaker take spoken commands, but no companion or conversation behavior is documented. |
| Specialized job | GoodSpecialized industrial and research tasks (assembly, sorting, gait training) are the core application domain. | GoodAn open end-effector connector and LeRobot-format datasets suit builders training one task of their own. |
| Supervision needs | ||
| Fully autonomous | PoorTeleoperation required for complex tasks; unattended autonomous operation is not supported. | PoorNo source shows the A3 finishing a chore unattended, and Nori reports no measured task performance. |
| With oversightthey differ | GoodDesigned for on-site technician supervision; teleoperation available as fallback for all tasks. | DependsTaught routines run through an off-board backend with no published success rates; each run needs watching. |
| Teleoperated | GoodIndustrial pilots with factory or research-facility integration are the primary deployment model. | GoodRecording demonstrations and iterating policies in Nori Lab is the workflow the platform is built around. |
| Timeline | ||
| Ready todaythey differ | PoorNot consumer-ready; institutional research and industrial pilots only. | DependsFirst units shipped July 21, 2026, but new orders ship fall 2026 with no published skill list. |
| Near-term (1–2 yrs)they differ | GoodMature institutional platform with 100+ deployments; consumer market timeline unknown. | DependsThe second batch ships fall 2026; the planned Skills Marketplace has no published launch date. |
| Research phase | GoodActive platform for academic locomotion and manipulation research at universities globally. | GoodNori's paper sets the A3 against research platforms, with datasets that plug into existing training pipelines. |
A robot being poor for a condition is a fact about fit, not a fault. “Depends” means the record names a limit worth reading, not that the answer is unknown.
How much the claim can be trusted.
None of these has shipped, so they are assessed on track record, engineering plausibility and how the demonstrations match the claims.
Track Record35% of score
Has this maker shipped what it demoed before, on schedule, with the promised capability?
8845
- Fourier GR-188
- Nori A345
28median 69 across 80 pre-release records100
Engineering35% of score
Are the demonstrated capabilities achievable in shipping form, or do they need breakthroughs not in evidence?
8368
- Fourier GR-183
- Nori A368
35median 70 across 87 pre-release records100
Demo Match30% of score
What has been shown to work, against what is claimed in marketing.
8838
- Fourier GR-188
- Nori A338
30median 66 across 87 pre-release records100
Published beside the number
Never averaged into it. These describe how much the maker has disclosed, not what the machine does.
| Fourier GR-1 | Nori A3 | |
|---|---|---|
| ReadinessShip-date specificity, pricing transparency, distribution and regulatory status. Published beside the number; never averaged in. | 50/100 | 74/100 |
| DisclosureHow much of the product is disclosed. Higher means fewer fundamental unknowns remain. Published beside the number; never averaged in. | 100/100 | 40/100 |
How far apart is a long way?
A number only means something next to the spread it sits in. Each track is the real range across every classified robot in the class, with the median marked.
Weight48 with a published figure
121.3 lbs45 lbs
- Fourier GR-1121.3 lbs
- Nori A345 lbs
26.5 lbsmedian 142.6 lbs661.4 lbs
Height51 with a published figure
65 in57.1 in
- Fourier GR-165 in
- Nori A357.1 in
34.7 inmedian 66.9 in86.6 in
Not compared, because fewer than two of these robots carry a published figure: Price, Runtime, Charge time, Noise.
What the numbers leave out.
Written on each record, and the only thing here every robot on the site carries.
The trade-off
What each design gives up to be good at what it is good at.
Fourier GR-1
GR-1 walks and reaches to a marked position on its own; any task beyond that defaults to a human teleoperator.
Documented deployments name Chinese hospitals and European auto plants as sites, not tasks: Fourier publishes no task-level detail for either program.
No published battery capacity or runtime figure exists for the GR-1, so planning a multi-day deployment depends on terms set by the distributor, not Fourier.
Fourier’s decade of rehab-robotics experience explains why the GR-1 was engineered for safe human proximity, but gait training and patient spotting belong to the ExoMotus exoskeleton line, not to any documented GR-1 task.
Nori A3
At $1,688 you get the hardware; the chores are a project you record yourself.
Off-board inference keeps the onboard computer at Raspberry Pi 5 cost, which makes the home network part of the robot.
Who each one is wrong for
Stated by the record itself. A robot being wrong for a household is a fact about fit, not a fault.
Fourier GR-1
- Consumers or small-business owners seeking autonomous help at home
- Fourier publishes no GR-1 price; distributors list it quote-only, with both purchase and lease paths. The platform requires on-site integration and dedicated technician availability, and no consumer support ecosystem exists.
- Environments without continuous supervision and safety protocols
- Teleoperation fallback is mandatory for complex tasks; unattended operation is not supported or recommended.
- Applications requiring instantaneous generalization across novel environments
- Operates within structured, pre-programmed task domains. Manual task scripting and environment reconfiguration required between deployments.
Nori A3
- Households wanting chores done on delivery
- No independent source documents which tasks work when the robot arrives, and Nori's own paper reports no measured task performance. Routines are recorded as demonstrations first.
- Multi-level homes needing help on more than one floor
- The base is wheeled and differentially driven. Nothing in Nori's documentation describes stair climbing, so a second floor means carrying a 20.4 kg robot.
- Homes with unreliable internet or no appetite for off-board processing
- Policy inference runs on an off-board backend, and Nori's paper names network connectivity as a requirement. Without it, the onboard Raspberry Pi 5 has no learned behavior to run.
- Tasks involving heavy items held in the grippers
- The 55 kg rating belongs to the lift column. Each arm is rated for 1.5 kg, which limits handling to light household objects.
- Buyers who need warranty terms and a support record
- Y Combinator lists Nori Robotics as founded in 2026 with five people, and the order page states no warranty or return terms.
What stays your job
The upkeep ownership hands back to you, whatever the autonomy level says.
Fourier GR-1
- Daily: Battery charge cycle check and dock connectivity verification.
- Weekly: Joint range-of-motion inspection on shoulders, elbows, hips, knees for actuator play or grinding noise.
- Monthly: Filter cleaning on cooling system (if deployed in dusty environments); sensor lens wipe on cameras and depth sensors.
- Every 3-6 months: Firmware update review and installation by technician.
- Annually: Full system calibration and load testing.
- As needed: Emergency manual override procedures if autonomous motion stalls.
- Service terms: Fourier publishes no lubrication interval, wear-replacement schedule or service pricing for the GR-1’s integrated FSA actuator modules. Those terms come from the supplying distributor and should be established before a deployment is budgeted.
Nori A3
- Daily: Recharge the battery, which Nori rates for 6 to 8 hours of operation.
- As needed: Record new demonstrations in Nori Lab to add or retrain a task.
- As needed: Restart work after the thermal interlock disables the motor bus at 138°F.
- As needed: Replace broken parts using the replacement CAD Nori supplies.
- As needed: Take backend policy updates, which reach deployed units without a hardware change.
- As needed: Clean the lidar window and four camera lenses when perception inputs look obstructed.
How much each of these is standing on.
Classifications rest on different depths of evidence. A pre-release assessment and a verified one are not the same kind of claim.
| Robot | Evidence status | Sources reviewed | Last assessed |
|---|---|---|---|
| Fourier GR-1 | Not Consumer-Ready | 5 | Aug 2026 |
| Nori A3 | Not Consumer-Ready | 8 | Sep 2026 |
Evidence status counts the sources behind a classification, not their kind. A high count can be maker documentation and price coverage, so read it as how much was reviewed rather than as independent testing. How we weigh evidence.
Keep going.
The full recordFourier GR-1Specifications, evidence, the Fit Check in full, and every tracked change.
The full recordNori A3Specifications, evidence, the Fit Check in full, and every tracked change.Robovations classifies robots; it does not rank them. Nothing on this page is a recommendation, and no manufacturer influences a classification or its placement here.
Add a robot
Comparing across classes is allowed — the page will switch to per-class positioning and tell you it has.