> ## Documentation Index
> Fetch the complete documentation index at: https://docs.novacula.io/llms.txt
> Use this file to discover all available pages before exploring further.

# Stable

> Run a Stable node — a CometBFT + EVM chain serving Ethereum JSON-RPC — Agent-only

Stable is a CometBFT-based chain with an EVM layer. A Stable node runs a single `stabled` process that serves both the **Ethereum JSON-RPC** and the **CometBFT RPC**, so it's modeled as one component rather than an execution/consensus split.

Stable runs **only on a bare-metal executor (Agent)** — it isn't offered on the Kubernetes (Operator) backend. To provision a host, see [Provision on bare-metal](/recipes/provision-on-bare-metal).

To deploy, open **Deploy Node**, pick **Stable**, then choose a network and a node type.

## Networks and resources

| Network               | Node type | CPU      | RAM    | Disk     |
| --------------------- | --------- | -------- | ------ | -------- |
| `mainnet` / `testnet` | Pruned    | 4 cores  | 8 GiB  | 200 GiB  |
| `mainnet` / `testnet` | Archive   | 16 cores | 32 GiB | 4096 GiB |

## Node types

* **Pruned** — keeps recent state and bootstraps fast from a published snapshot. Best for RPC serving.
* **Archive** — keeps all historical state; syncs from genesis (the published snapshot is pruned, so an archive node doesn't use it).

## Ports

| Name            | Port    | Notes                                            |
| --------------- | ------- | ------------------------------------------------ |
| JSON-RPC (HTTP) | `8545`  | Ethereum-compatible RPC — the endpoint apps call |
| WebSocket       | `8546`  | Ethereum-compatible WS                           |
| CometBFT RPC    | `26657` | Consensus RPC (internal)                         |
| gRPC            | `9090`  | Cosmos gRPC (internal)                           |
| P2P             | `26656` | Peer-to-peer                                     |

Ports are managed for you and aren't user-overridable.

## Snapshot bootstrap

A **Pruned** node seeds from a published snapshot on first deploy to cut sync time. An **Archive** node keeps every historical state, so it syncs from genesis instead of using the (pruned) snapshot.

## Configuration

Stable runs with a managed configuration — `config.toml`, `app.toml`, and the pinned genesis are rendered for you; there are no client config overrides at deploy time.

## Metrics

Block height, peers, and sync progress are surfaced in the node view — see [Node monitoring](/docs/nodes/monitoring).

## What you can build

* **Public RPC node** — accept the defaults (Pruned) and expose `:8545` (HTTP) / `:8546` (WS) behind your own proxy.
* **Archive node** — choose Archive for full historical state (heavier disk).

## Where to next

* [Deploy a node](/docs/nodes/deploy-a-node)
* [Provision on bare-metal](/recipes/provision-on-bare-metal)