pengjunjie/langflow
1
0
Fork 1
mirror of https://github.com/langflow-ai/langflow.git synced 2026-07-24 01:56:10 +08:00
Code Issues Packages Projects Releases Wiki Activity

docs: update about bundles page

Browse Source
This commit is contained in:
Mendon Kissling
2026-06-17 17:35:12 -04:00
parent b52185bc56
commit 175f3c4ad3
1 changed files with 74 additions and 87 deletions
Download Patch File Download Diff File Expand all files Collapse all files

161
docs/docs/Components/components-bundles.mdx
View File

@ -12,86 +12,20 @@ You add them to your flows and configure them in the same way as Langflow's core
To browse bundles, click <Icon name="Blocks" aria-hidden="true" /> **Bundles** in the visual editor.
## Install bundle components {#bundle-install}
## Core components and bundles
`uv pip install langflow` installs the Langflow web UI and a curated set of bundle components.
Many provider integrations ship as separate packages and are not installed by default.
:::tip
The Langflow documentation doesn't list all bundles or components in bundles.
For the most accurate and up-to-date list of bundles and components for your version of Langflow, check <Icon name="Blocks" aria-hidden="true" /> **Bundles** in the visual editor.
| Install command | What is installed |
|---|---|
| `uv pip install langflow` | Langflow web UI + core components + [select bundles](#bundles-included-by-default) |
| `uv pip install lfx-bundles` | All [long-tail provider bundles](#long-tail-bundles) (~45 providers), added to an existing Langflow install |
| `uv pip install "lfx[bundles]"` | The `lfx` engine + all long-tail bundles in one step — for [CLI-only use](#engine-only) |
| `uv pip install lfx-<provider>` | A [standalone partner bundle](#graduated-partner-bundles), such as `lfx-openai` |
| `uv pip install lfx` | The `lfx` engine only — no UI, no bundles. For [headless / CLI-only use](#engine-only). |
If you can't find a component that you used in an earlier version of Langflow, it may have been removed or marked as a [legacy component](#legacy-bundles).
:::
### Bundles included by default {#bundles-included-by-default}
Langflow offers generic <Icon name="Component" aria-hidden="true" /> **Core components** in addition to third-party, provider-specific bundles.
These bundles are included when you run `uv pip install langflow`:
If you are looking for a specific service or integration, you can <Icon name="Search" aria-hidden="true" /> **Search** components in the visual editor.
| Bundle | Package |
|---|---|
| DuckDuckGo | `lfx-duckduckgo` |
| Arxiv | `lfx-arxiv` |
| IBM | `lfx-ibm` |
| Docling | `lfx-docling` |
### Long-tail bundles {#long-tail-bundles}
About 45 provider bundles — including Chroma, Groq, Ollama, Mistral, Qdrant, and others — are distributed in the **`lfx-bundles`** metapackage.
**Already have Langflow installed?** Add just the bundle components:
```bash
uv pip install lfx-bundles
```
`langflow` already includes the `lfx` engine as a dependency, so installing `lfx-bundles` alone is enough. The engine discovers the new components automatically on the next restart.
**Starting from scratch with the `lfx` CLI?** Install the engine and all bundles together:
```bash
uv pip install "lfx[bundles]"
```
This is for environments where `langflow` is not installed and you only need the `lfx run` CLI to execute flows headlessly.
After installation, restart Langflow and confirm the bundles loaded:
```bash
lfx extension list
```
### Graduated partner bundles {#graduated-partner-bundles}
These flagship providers ship as standalone packages and must be installed individually:
| Provider | Package |
|---|---|
| OpenAI | `lfx-openai` |
| Anthropic | `lfx-anthropic` |
| Amazon | `lfx-amazon` |
| DataStax | `lfx-datastax` |
| Cohere | `lfx-cohere` |
For example:
```bash
uv pip install lfx-openai
```
After installation, restart Langflow.
### Install without bundles {#engine-only}
To install only the `lfx` execution engine — useful for headless deployments, CI pipelines, or embedding flow execution in another application without the web UI:
```bash
uv pip install lfx
```
This gives you the `lfx run` CLI for executing saved flows programmatically.
To add provider bundles later, see [Long-tail bundles](#long-tail-bundles) or [Graduated partner bundles](#graduated-partner-bundles) above.
If all else fails, you can always create your own [custom components](/components-custom-components).
## Bundle maintenance and documentation
@ -111,26 +45,79 @@ The Langflow documentation focuses on using bundles within flows.
For that reason, it focuses on the Langflow-specific configuration steps for bundled components.
For information about provider-specific features or APIs, see the provider's documentation.
## Component parameters
## Install bundle components {#bundle-install}
import PartialParams from '@site/docs/_partial-hidden-params.mdx';
Not all bundle components are included with `uv pip install langflow`.
<PartialParams />
| Bundle | Included with `uv pip install langflow`? | To install separately |
|---|---|---|
| DuckDuckGo, arXiv, IBM, Docling | Yes | — |
| Long-tail bundles (~45 providers) | No | `uv pip install lfx-bundles` |
| OpenAI, Anthropic, Amazon, Cohere | Core LLM support included; bundle adds enhanced parameters | `uv pip install lfx-<name>` |
| DataStax | No | `uv pip install lfx-datastax` |
## Core components and bundles
See the sections below for full package names and install details.
:::tip
The Langflow documentation doesn't list all bundles or components in bundles.
For the most accurate and up-to-date list of bundles and components for your version of Langflow, check <Icon name="Blocks" aria-hidden="true" /> **Bundles** in the visual editor.
If you can't find a component that you used in an earlier version of Langflow, it may have been removed or marked as a [legacy component](#legacy-bundles).
:::
### Bundles included by default {#bundles-included-by-default}
Langflow offers generic <Icon name="Component" aria-hidden="true" /> **Core components** in addition to third-party, provider-specific bundles.
The following bundles are included when you run `uv pip install langflow`:
If you are looking for a specific service or integration, you can <Icon name="Search" aria-hidden="true" /> **Search** components in the visual editor.
| Bundle | Package |
|---|---|
| DuckDuckGo | `lfx-duckduckgo` |
| Arxiv | `lfx-arxiv` |
| IBM | `lfx-ibm` |
| Docling | `lfx-docling` |
If all else fails, you can always create your own [custom components](/components-custom-components).
### Long-tail bundles {#long-tail-bundles}
About 45 provider bundles are distributed in the `lfx-bundles` metapackage.
If you already have Langflow installed, add the `lfx-bundles` package:
```bash
uv pip install lfx-bundles
```
`langflow` already includes the `lfx` engine as a dependency, so installing `lfx-bundles` alone is enough. The engine discovers the new components automatically on the next restart.
If you're starting from scratch with the `lfx` CLI, install the `lfx` engine and all bundles together:
```bash
uv pip install "lfx[bundles]"
```
This is for environments where `langflow` is not installed and you only need the `lfx run` CLI to execute flows headlessly.
After installation, restart Langflow and confirm the bundles loaded:
```bash
lfx extension list
```
### Graduated partner bundles {#graduated-partner-bundles}
These providers have standalone bundle packages with enhanced, provider-specific components.
For OpenAI, Anthropic, Amazon Bedrock, and Cohere, basic LLM support is already included in all Langflow installations through the core [Language Model](/components-models) component.
The bundle adds an enhanced component with additional provider-specific parameters.
To install a graduated bundle:
1. Run `uv pip install lfx-<name>` (for example, `uv pip install lfx-openai`).
2. Restart Langflow.
3. To confirm the bundle loaded, run `lfx extension list`.
### Install without bundles {#engine-only}
To install only the `lfx` execution engine — useful for headless deployments, CI pipelines, or embedding flow execution in another application without the web UI:
```bash
uv pip install lfx
```
This gives you the `lfx run` CLI for executing saved flows programmatically.
To add provider bundles later, see [Long-tail bundles](#long-tail-bundles) or [Graduated partner bundles](#graduated-partner-bundles) above.
## Legacy bundles