# Innate: an open robot platform for teachable physical agents

Canonical: https://mudpie.ai/companies/innate/
Breadcrumb: [Home](https://mudpie.ai/) / [Companies](https://mudpie.ai/companies/) / [Innate: an open robot platform for teachable physical agents](https://mudpie.ai/companies/innate/)
Author: Ali Abouelatta (https://mudpie.ai/authors/ali-abouelatta/)
Published: 2026-09-19
Updated: 2026-09-19
Research type: Company profile
Method: Company and accelerator sources checked 2026-09-19. Product claims are attributed to their sources; this is research, not a hands-on product trial.

Innate is selling a robot you teach, not a fixed home appliance.

## What it does

The current [Innate homepage](https://www.innate.bot/) positions MARS as a small, portable agentic robot. Its software layer, innateOS, is open source and lets builders simulate behaviors, write skills, teach tasks by demonstration, and run agents on the robot. The [documentation](https://docs.innate.bot/) says the software can run in a laptop simulator or on MARS, with Python skills, navigation, cameras, arm control and speech handled by the platform.

That makes the product easier to understand. Innate is not asking a buyer to buy a sealed robot and wait for a future app store. It is offering a hardware/software platform: write a skill, simulate it, try it on the robot, then share the behavior. The homepage lists examples including chess, sorting, operating a printer and cleaning a surface. Those are company-presented examples, not a measured reliability claim.

## Why I’d look closer

The price is visible. MARS is listed at $995 on the homepage and docs. The docs also list a Jetson Orin Nano, stereo depth, LiDAR, a gripper arm and an open-source software stack. That is enough information for a builder to decide whether the platform is worth experimenting with before asking about a larger deployment.

The founders’ public backgrounds fit the hardware/software ambition. The YC profile describes Axel Peytavin as a former Stanford HCI researcher and Ocean Cleanup team member. It describes Vignesh Anand as a former Stanford robotics graduate student and Archer flight-sciences intern who also built rockets and underwater robots. That is useful professional context, not proof that every hardware workflow is ready.

The tradeoff is the point. A platform this open asks the buyer to become part developer, part robot operator. Hardware availability, support, safety boundaries, accessory compatibility and the gap between simulation and a physical room matter more than a polished demo.

## My editorial take

Shortlist Innate for a builder who wants to teach and extend a physical agent. The simulator and open-source layer make the first experiment less risky than buying hardware blind. It is not the obvious fit for someone looking for a reliable household appliance with no code, calibration or tinkering. The product’s wedge is participation: users help decide what a general-purpose robot becomes.

## Quick facts

| Field | Sourced detail |
| --- | --- |
| Product | MARS robot plus open-source innateOS |
| Buyer | Robotics builders, researchers and technical early adopters |
| Price | MARS listed at $995 |
| Build path | Simulate → write or demonstrate a skill → run on MARS → share |
| Main question | How much of the simulator workflow transfers cleanly to a physical home environment? |

## Sources checked

| Source | Checked |
| --- | --- |
| [YC company profile](https://www.ycombinator.com/companies/innate) | 2026-09-19 |
| [Innate homepage](https://www.innate.bot/) | 2026-09-19 |
| [Innate documentation](https://docs.innate.bot/) | 2026-09-19 |
| [Innate about page](https://www.innate.bot/about) | 2026-09-19 |

## Cohort context

Innate is listed in Fall 2024. In our 2026-09-18 directory snapshot, 11 of 94 listed companies in that cohort have YC’s primary industry label Consumer (11.7%). This is a current-directory comparison, not an original intake count or a performance ranking. [Nine-cohort dataset](https://mudpie.ai/research/yc-cohorts-2026-09-19.json).

## Public website snapshot

Observed 2026-09-19T16:14:39.756Z in raw homepage HTML. This records visible metadata and advertised links, not agent execution or product quality.

| Signal | Homepage observation |
| --- | --- |
| Product description metadata | Observed |
| Canonical link | Observed |
| H1 or H2 heading | Observed |
| Typed structured data | Not observed in this response |
| Docs/developer link | Observed |
| Pricing link | Not observed in this response |
| llms.txt link | Not observed in this response |
| Markdown alternate | Not observed in this response |

[Public observations](https://mudpie.ai/research/yc-homepage-links-2026-09-19.json) · [Collection method](https://mudpie.ai/research/yc-homepage-methods/README.md). Missing links here do not establish that a capability or file is absent elsewhere.


## Author disclosure

I cofound Lazyweb and publish Mudpie. This is an owner-written publication, not an independent testing organization. Research notes distinguish observations, sourced reporting and editorial judgment.