# Pipeline Nodes

After a trigger, you build your pipeline by adding processing nodes in sequence. Each node performs one step and passes its output to the next. The pipeline editor only shows valid next nodes, so you always build a working chain.

### Node Types

#### Render

Creates a Glossi project for each model using a template, then renders images and videos. This node handles the full workflow — model preparation, project creation, and rendering — in one step.

**Configuration:**

* **Template** (required) — Select from your workspace templates. The template defines the scene — lighting, backdrops, camera angles, and environment.
* **Images** (toggle) — Render all camera bookmarks as still images.
  * **Quality:** 720p, 1080p, or 4K
* **Videos** (toggle) — Render all video shots and turntables.
  * **Resolution:** 720p, 1080p, or 4K
  * **Quality:** Fast (MP4) or Best (ProRes)
* **Render variants** (toggle) — If the model has variants (colors, materials), render each one.

For connector triggers (Watch Folder), the Render node also handles file ingestion automatically — downloading files from the connected storage, uploading them to Glossi, and creating 3D models before rendering.

**Multiple Render Nodes:** You can add multiple Render nodes to a single pipeline, each with a different template. This is useful when you need the same model rendered in multiple scenes (e.g., "Studio White" and "Lifestyle Kitchen"). Each Render node creates its own project and processes independently. Click the "+" button on the trigger node to add sibling Render nodes.

**Follows:** Any trigger (Watch Folder, Model Ready, Project Created) **Next:** Style, AI Edit, or Export

***

#### Style (Apply Brand Styles)

Generates styled image variants using your workspace's [Brand Styles](https://docs.glossi.io/workspace/brand-styles). Each selected style creates one variant per rendered image.

For example: 3 camera bookmarks × 2 styles = 6 styled variants (plus the 3 original renders).

If a style variant fails (e.g., AI service timeout), the pipeline automatically retries it up to 2 times before moving on. Only permanently failed variants are skipped — successful variants still pass to downstream nodes.

**Configuration:**

* **Styles** (required) — Select one or more Brand Styles from the grid.
* **Prompt** (optional) — Guide the style application with a text description.
* **Group by style** — Controls how styled variants are organized in the export folder:
  * **Off (default):** Styled variants sit alongside base renders with the style name appended (e.g., `Front View_Warm Studio.png`).
  * **On:** Styled variants are organized into style-named subfolders (e.g., `Warm Studio/Front View.png`).

You can chain multiple Style nodes. Each node receives the output of the previous one, allowing layered styling.

**Follows:** Render, Render Complete **Next:** Style (chain), AI Edit, or Export

***

#### AI Edit (AI Variant)

Creates AI-edited image variants using text prompts. Each prompt generates a separate variant for every rendered image.

If an AI edit fails, the pipeline automatically retries it up to 2 times before moving on.

**Configuration:**

* **Prompts** — Add one or more text prompts. Each prompt creates a distinct variant.
  * Example: "Place the product on a marble countertop with soft morning light"
  * Example: "Add a gradient background transitioning from blue to purple"

**Follows:** Render, Style, Render Complete **Next:** Export (terminal — no further nodes)

***

#### Export

Uploads completed renders (and any styled or AI-edited variants) to a connected storage folder.

**Configuration:**

* **Account** — Sign in with OAuth, or use the same account as the trigger connector if it's the same provider.
* **Folder** — Browse and select the destination folder.

When the trigger and export use the same connector type (e.g., both Google Drive), you can click **"Use \[account]"** to share credentials without signing in again. You can also sign in with a different account.

**Follows:** Render, Style, AI Edit **Next:** None — Export is always the last node.

### Valid Node Sequences

The pipeline editor enforces valid transitions. Here's what can follow each node:

| After this node... | You can add...            |
| ------------------ | ------------------------- |
| Watch Folder       | Render                    |
| Model Ready        | Render                    |
| Project Created    | Render                    |
| Render Complete    | Style, AI Edit, or Export |
| Render             | Style, AI Edit, or Export |
| Style              | Style, AI Edit, or Export |
| AI Edit            | Export (terminal)         |

### Example Pipelines

#### Basic: Google Drive to Renders

```
Watch Folder (Google Drive) → Render → Export (Google Drive)
```

#### With Brand Styles

```
Watch Folder (OneDrive) → Render → Style → Export (OneDrive)
```

#### Chained Styles

```
Watch Folder (Google Drive) → Render → Style (Brand Style A) → Style (Brand Style B) → Export
```

#### Event-Driven Styling

```
Render Complete → Style
```

Applies Brand Styles to every render in your workspace automatically.

#### AI Variants on New Renders

```
Render Complete → AI Edit
```

Creates AI-edited variants using your prompts whenever a render completes.

#### Multi-Template Rendering

```
Watch Folder (Google Drive) → Render (Studio White) → Export (Google Drive)
                            └→ Render (Lifestyle Kitchen) → Style → Export (Google Drive)
                            └→ Render (Outdoor Scene) → Export (Google Drive)
```

Each model is rendered with three different templates. Each Render branch can have its own variant and export nodes.