### [Open WebUI Pipelines](https://github.com/open-webui/pipelines)
> Handle: `pipelines`
> URL: [http://localhost:34211](http://localhost:34211)
> Integration author: [@ic4l4s9c](https://github.com/ic4l4s9c)
Pipelines bring modular, customizable workflows to any UI client supporting OpenAI API specs – and much more! Easily extend functionalities, integrate unique logic, and create dynamic workflows with just a few lines of code.
---
#### Starting
```bash
# [Optional] Pre-pull the images
harbor pull pipelines
# Start the service
harbor up pipelines
```
> [!NOTE]
> It makes most sense to run `pipelines` and `webui` together, as they are designed to work in tandem
#### Usage
Harbor connects `pipelines` to the `webui` service by default, when run together.

Apart from that, you can access the Pipelines API:
```bash
# URL for access from the host
harbor url pipelines
# URL for access in the same LAN
harbor url -a pipelines
# [Optional] create tunnel for access via
# public interned
# ⚠️ Never expose unprotected endpoints to the public internet
harbor tunnel pipelines
```
#### Configuration
Majority of configuration will happen via `webui`. Navigate to the pipelines settings (typically `Admin Settings` -> `Settings` -> `Pipelines`) and upload either a `.py` file with the pipeline code or point the service to a _raw_ GitHub URL (for example: https://raw.githubusercontent.com/open-webui/pipelines/refs/heads/main/examples/pipelines/integrations/wikipedia_pipeline.py)
After upload, the usage of the pipeline in the `webui` will depend on the type of the pipeline (e.g. Filter, Function, Scaffold, etc.).

Please refer to the [official documentation](https://docs.openwebui.com/pipelines/) for many more examples and detailed instructions.
#### Documentation
- [Official documentation](https://docs.openwebui.com/pipelines/)
- [Examples on GitHub](https://github.com/open-webui/pipelines/tree/main/examples)