Cyborg Index · August 2026
Figure 03 vs Kepler Forerunner K3
Both humanoids read on the same 8 axes and 40 sub-axes, using the published Index scores. Every difference below carries the public evidence behind it, and sub-axes with no public evidence count as 0 rather than being skipped.
Overall standing
Capability score · 0–5 over 40 sub-axesThe short answer
Figure 03 and Kepler Forerunner K3 score the same — 1.20 and 1.18.
Nothing separates them overall. The table below shows the axes where they actually differ.
Where Figure 03 is ahead
- Safety+0.80Rests on claims and demos
- Learning+0.20Rests on claims and demos
- Deployment+0.20Rests on claims and demos
Where Kepler Forerunner K3 is ahead
- Locomotion+0.80Rests on claims and demos
- Manipulation+0.20Rests on claims and demos
What would change it
6 sub-axes on the trailing side have no public evidence yet. If all of them scored full marks they would be worth 0.75 — against a gap of 0.02.
- Learning · Data flywheelup to +0.125
- Perception · Spatial memoryup to +0.125
- Autonomy · Generalizationup to +0.125
- Locomotion · Uneven terrainup to +0.125
- Locomotion · Disturbance recoveryup to +0.125
How much we know
68%of sub-axes covered
We have public evidence for 68% of the 40 sub-axes across these two.
Figure 03 · 68%
Kepler Forerunner K3 · 68%
The shape of each score
27/40 sub-axes have evidence
Faint outline = Kepler Forerunner K3
27/40 sub-axes have evidence
Faint outline = Figure 03
Reading the wheels — clockwise from the top
- 1 Locomotion
- 2 Manipulation
- 3 Perception
- 4 Autonomy
- 5 Learning
- 6 HRI
- 7 Deployment
- 8 Safety
Each wheel has 40 spokes — five per axis. A longer spoke is a higher score, and the dashed rings mark 1 to 4. Hover or tab onto any spoke to read the axis, the sub-axis and the exact value. A missing spoke means there is no public evidence for that sub-axis yet, which is not the same as scoring zero. The faint outline on each wheel is the other humanoid, drawn on the same scale.
Every axis, every sub-axis
Figure 03Kepler Forerunner K3
Shared 0–5 track · widest gap firstLocomotion
0.80 / 1.60► 0.80
Rests on claims and demosEvidenced cells · 2/5 vs 4/5
Flat ground walking
2.00 / 2.00
Uneven terrain
— / 2.00► 2.00
Stairs
— / —
Disturbance recovery
— / 2.00► 2.00
Speed endurance
2.00 / 2.00
Safety
1.00 / 0.20◄ 0.80
Rests on claims and demosEvidenced cells · 3/5 vs 1/5
Functional safety
2.00 / 1.00◄ 1.00
Certifications
2.00 / —◄ 2.00
E stop arch
— / —
Incident transparency
— / —
Force impact limits
1.00 / —◄ 1.00
Manipulation
1.00 / 1.20► 0.20
Rests on claims and demosEvidenced cells · 3/5 vs 3/5
Grasp variety
2.00 / 2.00
Bimanual
1.00 / 2.00► 1.00
Force control
2.00 / 2.00
In hand reorient
— / —
Tool use
— / —
Learning
1.40 / 1.20◄ 0.20
Rests on claims and demosEvidenced cells · 4/5 vs 3/5
Imitation learning
1.00 / 2.00► 1.00
Sim to real
2.00 / —◄ 2.00
Foundation policy
2.00 / 2.00
Data flywheel
— / 2.00► 2.00
Transfer fine tune
2.00 / —◄ 2.00
Deployment
1.20 / 1.00◄ 0.20
Rests on claims and demosEvidenced cells · 4/5 vs 4/5
Named pilots
2.00 / 1.00◄ 1.00
Field hours
1.00 / 1.00
Customer diversity
1.00 / 2.00► 1.00
Supportability
2.00 / —◄ 2.00
Commercial posture
— / 1.00► 1.00
Perception
1.40 / 1.40
Not enough public record yetEvidenced cells · 4/5 vs 4/5
Sensor stack
2.00 / 2.00
Scene understanding
1.00 / 1.00
Spatial memory
— / 2.00► 2.00
Object permanence
2.00 / —◄ 2.00
Semantic world model
2.00 / 2.00
Autonomy
1.20 / 1.20
Not enough public record yetEvidenced cells · 3/5 vs 4/5
Unattended operation
2.00 / 0.00◄ 2.00
Task horizon
2.00 / 2.00
Generalization
— / 2.00► 2.00
Instruction following
2.00 / 2.00
Recovery from failure
— / —
HRI
1.60 / 1.60
Not enough public record yetEvidenced cells · 4/5 vs 4/5
Speech io
2.00 / 2.00
Nl instruction
2.00 / 2.00
Intent recognition
— / —
Safe proximity
2.00 / 2.00
Social presentation
2.00 / 2.00
Segment length and thickness are the gapNo public evidence yet
Which is better for what
Each row is decided by the axes that matter for that kind of work, read off the same Capability scores as the ranking. Rows only appear where both humanoids have enough public record to answer.
Manufacturing and warehouse work
Named pilots and field hours, backed by what the hands can actually do.
Level
Figure 03 and Kepler Forerunner K3 read the same on Deployment and Manipulation (1.13 vs 1.07).
Running without a human in the loop
Unattended operation and task horizon, backed by what the robot can perceive.
Level
Figure 03 and Kepler Forerunner K3 read the same on Autonomy and Perception (1.27 vs 1.27).
Moving through difficult ground
Uneven terrain, stairs and disturbance recovery.
Kepler Forerunner K3
Kepler Forerunner K3 scores higher on Locomotion — 1.60 against 0.80 for Figure 03.
Working alongside people
Functional safety and certification, plus how it reads and answers people.
Figure 03
Figure 03 scores higher on Safety and HRI — 1.20 against 0.67 for Kepler Forerunner K3.
Picking up new tasks
Imitation learning, transfer and generalisation across tasks.
Figure 03
Figure 03 scores higher on Learning and Autonomy — 1.33 against 1.20 for Kepler Forerunner K3.
Research and developer use
How much of the stack is documented and reproducible.
Figure 03
Figure 03 scores higher on Learning and Perception — 1.40 against 1.27 for Kepler Forerunner K3.
How solid the record is on each side
Figure 03
- Sub-axes on the record
- 27 of 40 · 68%
- Carried by verified material
- 8%
- Rests on claims and demos
- 1.10
Best documented — Manipulation, Perception, Autonomy, Learning, HRI, Deployment, Safety
Nothing on the record — every axis has at least one read
Kepler Forerunner K3
- Sub-axes on the record
- 27 of 40 · 68%
- Carried by verified material
- 6%
- Rests on claims and demos
- 1.10
Best documented — Locomotion, Manipulation, Perception, Autonomy, Learning, HRI, Deployment
Nothing on the record — every axis has at least one read
What this comparison cannot settle
- Neither humanoid has a public read on 7 of the 40 sub-axes, so those cannot be compared at all.
- 12 sub-axes are read on only one side. A one-sided cell is undecided, not a zero for the silent side.
- Most of Figure 03's score rests on vendor statements and demo footage rather than independently verified material (8% carried by verified evidence).
- Most of Kepler Forerunner K3's score rests on vendor statements and demo footage rather than independently verified material (6% carried by verified evidence).
- Demonstrations happen in different environments and product revisions are not always equivalent, so like-for-like uptime and throughput figures do not exist for most of the field.