What THOR is

THOR is building open robotics infrastructure, starting with a demand-first marketplace for commissioning training data. Expert robotics annotation is the first product: customer contracts define the work, qualified contributors produce it and THOR coordinates verification and delivery.

The economic model connects that demand to SN112: treasury-held alpha funds reserved worker rewards, while customer receipts allocated to buybacks fund purchases of the subnet's alpha token. Demand-first token economics explains the flows and why the first phase is a private pilot.

THOR Data develops that market. Its longer-term ambition is to serve developers and AI agents commissioning complete datasets, beginning with robotics and expanding through demonstrated demand. The same commissioning and verification methods could serve other learning tasks alongside THOR Group's wider Data → Models → Hardware direction.

Data, Simulation, Hardware, Research, China and Deployment address connected parts of that robotics mission. SN112 is THOR's subnet on Bittensor, intended to link useful contribution with network incentives. Why Bittensor distinguishes the group, the service and the network.

The wider ecosystem

The THOR ecosystem maps the six areas and explains their roles and development focus.

THOR Data is the starting point: qualified firms fulfil buyer-defined briefs during the private beta, followed by permissionless annotation after beta. Research, simulation, hardware, manufacturing relationships and deployment extend the wider mission around demonstrated needs.

Commercial work and shared progress

Commercial demand can help sustain open research resources, models and tools as complementary services develop. Shared progress comes from authorised releases, reusable methods and delivery experience, while customer material remains governed by its agreed rights.

How the work compounds

The connections matter: research informs data requirements, simulation supports evaluation, and hardware and deployment generate new physical experience. A failure observed in the field can become a new research question or production brief.

A completed project can also leave a sharper specification, reusable tooling or evidence that improves evaluation. Where rights permit, reviewed examples and documented errors can refine future checks. This is the compounding model THOR aims to build; it depends on reliable records and demonstrated improvements, while respecting customer confidentiality.