> For the complete documentation index, see [llms.txt](https://deepnode.gitbook.io/docs/llms.txt). Markdown versions of documentation pages are available by appending `.md` to page URLs; this page is available as [Markdown](https://deepnode.gitbook.io/docs/about-us/technical-faqs/roles.md).

# Roles

<details>

<summary>What are the main roles in the DeepNode ecosystem?</summary>

In Phase 1, DeepNode includes six key roles:

* **Model Creators** who upload and monetize AI models.
* **Validators** who validate and score models for quality.
* **Miners** who provide compute power for inference.
* **Stakers** who delegate $DN to nodes, securing the protocol.
* **Backers** who support models directly in exchange for revenue shares.
* **Consumers** who use AI services and drive demand.

In Phase 2, **Domain Architects** join, launching and governing domain-specific AI economies.

In Phase 3, **Data Providers** and **Data Annotators** expand the ecosystem into the data layer.

</details>

<details>

<summary>What are the responsibilities, rewards, and requirements for each role?</summary>

**Model Creators**

* *Responsibilities:* Upload, register, and maintain AI models.
* *Rewards:* Earn $DN per model usage and support from Backers.
* *Requirements:* AI/ML expertise and compliance with validation standards.

**Validators**

* *Responsibilities:* Validate models, assign trust scores, ensure quality.
* *Rewards:* Earn $DN for accurate evaluations, influenced by Dynamic Trust Weights.
* *Requirements:* Analytical expertise and optional compute resources.

**Miners**

* *Responsibilities:* Run the Node Execution Engine and provide reliable inference.
* *Rewards:* Earn $DN from task execution, boosted by Stakers’ support.
* *Requirements:* GPU or CPU hardware and technical operational capacity.

**Stakers**

* *Responsibilities:* Delegate $DN to trusted nodes or Validators and monitor performance.
* *Rewards:* Earn staking rewards and gain governance power.
* *Requirements:* Hold $DN and evaluate operator performance.

**Backers**

* *Responsibilities:* Support promising models by bonding $DN.
* *Rewards:* Revenue shares proportional to model adoption.
* *Requirements:* Hold $DN and assess model potential and Validator scores.

**Consumers**

* *Responsibilities:* Use models via the portal or APIs, pay with $DN, and provide feedback.
* *Rewards:* Access diverse, validated AI services at competitive cost.
* *Requirements:* Hold $DN and choose models based on ratings.

**Domain Architects (Phase 2)**

* *Responsibilities:* Launch domains, set rules, and coordinate operators and Validators.
* *Rewards:* Earn fees and reputation from subnet activity.
* *Requirements:* Technical, governance, and economic design skills.

**Data Providers (Phase 3)**

* *Responsibilities:* Supply raw datasets for training and evaluation.
* *Rewards:* Earn $DN when datasets are used.
* *Requirements:* Data ownership rights and formatting capability.

**Data Annotators (Phase 3)**

* *Responsibilities:* Label and structure datasets to improve quality.
* *Rewards:* Earn $DN for quality annotations.
* *Requirements:* Annotation expertise and accuracy.

</details>

<details>

<summary>What environments are supported for running different roles?</summary>

DeepNode supports a wide range of operating environments so participants can choose setups that match their scale and expertise:

* **Bare metal servers** — Ideal for Miners needing direct access to high-performance GPUs or CPUs, with maximum control over efficiency and uptime.
* **Virtual machines (VMs)** — Common for Validators and Miners, offering isolation and resource guarantees within enterprise or data center clusters. This is useful when separating roles (Validator in one VM, Miner in another).
* **Cloud infrastructure** — Supported across all roles, including Model Creators (for model hosting), Validators, and Miners. Platforms like AWS, GCP, and Azure provide on-demand scaling, while decentralized cloud providers offer trust-minimized setups.

</details>

<details>

<summary>Can roles be combined or must they be isolated?</summary>

Roles in DeepNode are modular, and combinations are allowed depending on participant's resources and risk profile:

* **Model Creators** can also operate as **Miners** if they have compute power. This is a common dual role, they can mine, mint, and run their own models. However, their models will still need to pass through validation like any other.
* **Validators** can technically combine with **Miners**, but they cannot validate their own models. The protocol enforces separation of validation tasks to maintain fairness and avoid conflicts of interest.
* **Stakers** and **Backers** are token-based roles that can be freely combined with any other role.
* **Consumers** naturally overlap with all roles since everyone in the ecosystem ultimately uses AI services.
* **Domain Architects** may hold multiple roles, such as running nodes within their subnet, but governance encourages separation where possible to preserve neutrality.

In practice, the most common dual role is **Creator + Miner**, while **Evaluator** roles are intentionally isolated when interacting with their own models to preserve trust.

</details>