feat: Add Tzafon AI Integration

src/oss/javascript/integrations/chat/tzafon.mdx

@@ -0,0 +1,73 @@
+---
+title: ChatTzafon
+---
+
+[Tzafon](https://www.tzafon.ai/) is a research firm building scalable compute systems and foundation models for real-world interaction. The `ChatTzafon` class provides access to Tzafon's OpenAI-compatible [Chat Completions API](https://docs.tzafon.ai/core-concepts/chat-completions).
+
+## Setup
+
+### Installation
+
+Install the integration package:
+
+<CodeGroup>
+```bash npm
+npm install @tzafon/langchain-tzafon
+```
+```bash pnpm
+pnpm add @tzafon/langchain-tzafon
+```
+</CodeGroup>
+
+### Credentials
+
+Get an API key from [tzafon.ai](https://www.tzafon.ai/) and set it as an environment variable:
+
+```bash
+export TZAFON_API_KEY="your-api-key"
+```
+
+## Instantiation
+
+```typescript
+import { ChatTzafon } from "@tzafon/langchain-tzafon";
+
+const model = new ChatTzafon({
+  model: "tzafon.sm-1",
+  temperature: 0,
+  maxTokens: 1024,
+  apiKey: process.env.TZAFON_API_KEY,
+});
+```
+
+## Invocation
+
+```typescript
+import { HumanMessage } from "@langchain/core/messages";
+
+const messages = [new HumanMessage("Hello, how are you?")];
+const response = await model.invoke(messages);
+console.log(response.content);
+```
+
+## Streaming
+
+```typescript
+const stream = await model.stream(messages);
+for await (const chunk of stream) {
+  process.stdout.write(chunk.content);
+}
+```
+
+## Available models
+
+| Model | Description |
+|-------|-------------|
+| `tzafon.sm-1` | General-purpose chat model |
+| `tzafon.northstar.cua.sft` | Model trained for web navigation |
+
+Use the [List Models API](https://docs.tzafon.ai/core-concepts/chat-completions#list-models) to get the latest available models.
+
+## API reference
+
+For detailed API documentation, see the [Tzafon SDK documentation](https://docs.tzafon.ai/core-concepts/chat-completions).

src/oss/javascript/integrations/document_loaders/index.mdx

@@ -83,6 +83,7 @@ LangChain.js categorizes document loaders in two different ways:
 | [`Spider`](/oss/integrations/document_loaders/web_loaders/spider) | Fast crawler that converts websites into HTML, markdown, or text | ✅ | API |
 | [`RecursiveUrlLoader`](/oss/integrations/document_loaders/web_loaders/recursive_url_loader) | Recursively load webpages following links | ❌ | Package |
 | [`Sitemap`](/oss/integrations/document_loaders/web_loaders/sitemap) | Load all pages from a sitemap.xml | ✅ | Package |
+| [`Tzafon`](/oss/integrations/document_loaders/web_loaders/tzafon) | Programmatic control of browsers and desktops in seconds. | ✅ | Package |
 | [`Browserbase`](/oss/integrations/document_loaders/web_loaders/browserbase) | Load webpages using managed headless browsers with stealth mode | ✅ | API |
 | [`WebPDFLoader`](/oss/integrations/document_loaders/web_loaders/pdf) | Load PDF files in web environments | ✅ | Package |
 
@@ -188,6 +189,7 @@ LangChain.js categorizes document loaders in two different ways:
 <Card title="Subtitles" icon="link" href="/oss/integrations/document_loaders/file_loaders/subtitles" arrow="true" cta="View guide" />
 <Card title="Taskade" icon="link" href="/oss/integrations/document_loaders/web_loaders/taskade" arrow="true" cta="View guide" />
 <Card title="Text" icon="link" href="/oss/integrations/document_loaders/file_loaders/text" arrow="true" cta="View guide" />
+<Card title="Tzafon" icon="link" href="/oss/integrations/document_loaders/web_loaders/tzafon" arrow="true" cta="View guide" />
 <Card title="UnstructuredLoader" icon="link" href="/oss/integrations/document_loaders/file_loaders/unstructured" arrow="true" cta="View guide" />
 <Card title="WebPDFLoader" icon="link" href="/oss/integrations/document_loaders/web_loaders/pdf" arrow="true" cta="View guide" />
 <Card title="YouTube" icon="link" href="/oss/integrations/document_loaders/web_loaders/youtube" arrow="true" cta="View guide" />

src/oss/javascript/integrations/document_loaders/web_loaders/index.mdx

@@ -247,6 +247,13 @@ These loaders are used to load web resources. They do not involve the local file
     arrow="true"
     cta="View guide"
   />
+  <Card
+    title="Tzafon Loader"
+    icon="link"
+    href="/oss/integrations/document_loaders/web_loaders/tzafon"
+    arrow="true"
+    cta="View guide"
+  />
   <Card
     title="YouTube transcripts"
     icon="link"

src/oss/javascript/integrations/document_loaders/web_loaders/tzafon.mdx

@@ -0,0 +1,49 @@
+---
+title: Tzafon Loader
+sidebarTitle: Tzafon Loader
+---
+
+## Description
+
+[Tzafon](https://docs.tzafon.ai/overview) programmatic control of browsers and desktops in seconds. Full stealth. Lightning fast.
+
+## Installation
+
+Install the necessary packages:
+
+<CodeGroup>
+```bash npm
+npm install langchain-tzafon @langchain/core tzafon
+```
+```bash pnpm
+pnpm add langchain-tzafon @langchain/core tzafon
+```
+```bash yarn
+yarn add langchain-tzafon @langchain/core tzafon
+```
+```bash bun
+bun add langchain-tzafon @langchain/core tzafon
+```
+
+</CodeGroup>
+
+## Example
+
+Utilize the `TzafonLoader` as follows to allow your agent to load websites:
+
+```typescript
+import { TzafonLoader } from "langchain-tzafon";
+
+const loader = new TzafonLoader(["https://example.com"], {
+    apiKey: "your_api_key", // Optional if TZAFON_API_KEY is set in environment
+});
+const documents = await loader.load();
+```
+
+## Arguments
+
+-   `urls`: Required. List of URLs to load.
+
+## Options
+
+-   `apiKey`: Optional. Tzafon API key. Default is `TZAFON_API_KEY` env variable.

src/oss/javascript/integrations/providers/all_providers.mdx

@@ -1524,6 +1524,13 @@ Connect LangGraph agents to front ends.
     Load projects and tasks from Taskade.
   </Card>
 
+  <Card
+    title="Tzafon"
+    href="/oss/integrations/document_loaders/web_loaders/tzafon"
+  >
+    Programmatic control of browsers and desktops in seconds.
+  </Card>
+
   <Card
     title="Web Cheerio"
     href="/oss/integrations/document_loaders/web_loaders/web_cheerio"

src/oss/javascript/integrations/providers/tzafon.mdx

@@ -0,0 +1,47 @@
+---
+title: Tzafon
+---
+
+This page covers all LangChain.js integrations with [Tzafon](https://www.tzafon.ai/), a research firm building scalable compute systems and foundation models for real-world interaction.
+
+Tzafon provides an OpenAI-compatible [Chat Completions API](https://docs.tzafon.ai/core-concepts/chat-completions) with models like `tzafon.sm-1` and `tzafon.northstar.cua.sft`.
+
+## Installation and setup
+
+1. Get an API key from [tzafon.ai](https://www.tzafon.ai/) and set it as the environment variable `TZAFON_API_KEY`.
+2. Install the integration package:
+
+<CodeGroup>
+```bash npm
+npm install @tzafon/langchain-tzafon
+```
+```bash pnpm
+pnpm add @tzafon/langchain-tzafon
+```
+</CodeGroup>
+
+## Chat model
+
+See the [ChatTzafon usage example](/oss/integrations/chat/tzafon).
+
+```typescript
+import { ChatTzafon } from "@tzafon/langchain-tzafon";
+
+const model = new ChatTzafon({
+  model: "tzafon.sm-1",
+  apiKey: process.env.TZAFON_API_KEY,
+});
+```
+
+## Available models
+
+| Model | Description |
+|-------|-------------|
+| `tzafon.sm-1` | General-purpose chat model |
+| `tzafon.northstar.cua.sft` | Model trained for web navigation |
+
+Use the [List Models API](https://docs.tzafon.ai/core-concepts/chat-completions#list-models) to get the latest available models.
+
+## Document loader
+
+See the [TzafonLoader usage example](/oss/integrations/document_loaders/web_loaders/tzafon).

src/oss/python/integrations/chat/tzafon.mdx

@@ -0,0 +1,80 @@
+---
+title: ChatTzafon
+---
+
+[Tzafon](https://www.tzafon.ai/) is a research firm building scalable compute systems and foundation models for real-world interaction. The `ChatTzafon` class provides access to Tzafon's OpenAI-compatible [Chat Completions API](https://docs.tzafon.ai/core-concepts/chat-completions).
+
+## Setup
+
+### Installation
+
+Install the integration package:
+
+<CodeGroup>
+```bash pip
+pip install -U langchain-tzafon
+```
+```bash uv
+uv add langchain-tzafon
+```
+</CodeGroup>
+
+### Credentials
+
+Get an API key from [tzafon.ai](https://www.tzafon.ai/) and set it as an environment variable:
+
+```bash
+export TZAFON_API_KEY="your-api-key"
+```
+
+## Instantiation
+
+```python
+import os
+from langchain_tzafon import ChatTzafon
+
+os.environ["TZAFON_API_KEY"] = "..."
+
+model = ChatTzafon(
+    model="tzafon.sm-1",
+    temperature=0,
+    max_tokens=1024,
+)
+```
+
+
+## Invocation
+
+```python
+from langchain.messages import HumanMessage
+
+messages = [HumanMessage(content="Hello, how are you?")]
+response = model.invoke(messages)
+print(response.content)
+```
+
+## Streaming
+
+```python
+for chunk in model.stream(messages):
+    print(chunk.content, end="", flush=True)
+```
+
+## Async
+
+```python
+response = await model.ainvoke(messages)
+```
+
+## Available models
+
+| Model | Description |
+|-------|-------------|
+| `tzafon.sm-1` | General-purpose chat model |
+| `tzafon.northstar.cua.sft` | Model trained for web navigation |
+
+Use the [List Models API](https://docs.tzafon.ai/core-concepts/chat-completions#list-models) to get the latest available models.
+
+## API reference
+
+For detailed API documentation, see the [Tzafon SDK documentation](https://docs.tzafon.ai/core-concepts/chat-completions).

src/oss/python/integrations/document_loaders/index.mdx

@@ -316,6 +316,7 @@ The below document loaders allow you to load data from common data formats.
 <Card title="Trello" icon="link" href="/oss/integrations/document_loaders/trello" arrow="true" cta="View guide" />
 <Card title="TSV" icon="link" href="/oss/integrations/document_loaders/tsv" arrow="true" cta="View guide" />
 <Card title="Twitter" icon="link" href="/oss/integrations/document_loaders/twitter" arrow="true" cta="View guide" />
+<Card title="Tzafon" icon="link" href="/oss/integrations/document_loaders/tzafon" arrow="true" cta="View guide" />
 <Card title="UnDatasIO" icon="link" href="/oss/integrations/document_loaders/undatasio" arrow="true" cta="View guide" />
 <Card title="Unstructured" icon="link" href="/oss/integrations/document_loaders/unstructured_file" arrow="true" cta="View guide" />
 <Card title="UnstructuredMarkdownLoader" icon="link" href="/oss/integrations/document_loaders/unstructured_markdown" arrow="true" cta="View guide" />

src/oss/python/integrations/document_loaders/tzafon.mdx

@@ -0,0 +1,40 @@
+---
+title: Tzafon
+sidebarTitle: Tzafon Loader
+---
+
+## Description
+
+[Tzafon](https://docs.tzafon.ai/overview) programmatic control of browsers and desktops in seconds. Full stealth. Lightning fast.
+
+## Installation and setup
+
+Install the integration package:
+
+<CodeGroup>
+```bash pip
+pip install langchain-tzafon
+```
+```bash uv
+uv add langchain-tzafon
+```
+</CodeGroup>
+
+## Loading documents
+
+You can load webpages into LangChain using `TzafonLoader`.
+
+```python
+from langchain_tzafon import TzafonLoader
+
+# Ensure TZAFON_API_KEY is set in your environment variables, or pass it directly.
+loader = TzafonLoader(urls=["https://example.com"], api_key="your_api_key")
+documents = loader.load()
+print(documents[0].page_content[:10])
+```
+
+### Loader Options
+
+-   `urls` Required. A list of URLs to fetch.
+-   `api_key` Optional. Tzafon API key. If not provided, it will look for `TZAFON_API_KEY` in environment variables.
+-   `text_content` Optional. Retrieve only text content. Default is True.

src/oss/python/integrations/providers/all_providers.mdx

@@ -2951,6 +2951,14 @@ Browse the complete collection of integrations available for Python. LangChain P
         Fast, cost-efficient vector database for search and retrieval.
     </Card>
 
+    <Card
+        title="Tzafon"
+        href="/oss/integrations/providers/tzafon"
+        icon="link"
+    >
+        Programmatic control of browsers and desktops in seconds.
+    </Card>
+
     <Card
         title="UnDatasIO"
         href="/oss/integrations/providers/undatasio"