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

# Together AI

> Learn how to build AI agents that integrate Together AI's powerful LLMs with Klavis MCP Servers

## Prerequisites

Before we begin, you'll need:

<CardGroup cols={2}>
  <Card title="Together AI API Key" icon="key" href="https://together.ai/">
    Get your API key from Together AI
  </Card>

  <Card title="Klavis AI API Key" icon="key" href="https://klavis.ai/">
    Get your API key from Klavis AI
  </Card>
</CardGroup>

## Installation

First, install the required packages:

<CodeGroup>
  ```bash Python theme={null}
  pip install together klavis
  ```

  ```bash TypeScript theme={null}
  npm install together-ai klavis
  ```
</CodeGroup>

## Setup Environment Variables

<CodeGroup>
  ```python Python theme={null}
  import os

  # Set environment variables
  os.environ["TOGETHER_API_KEY"] = "your-together-api-key-here"  # Replace with your actual Together API key
  os.environ["KLAVIS_API_KEY"] = "your-klavis-api-key-here"      # Replace with your actual Klavis API key
  ```

  ```typescript TypeScript theme={null}
  // Set environment variables in your .env file
  process.env.TOGETHER_API_KEY = "your-together-api-key-here";  // Replace with your actual Together API key
  process.env.KLAVIS_API_KEY = "your-klavis-api-key-here";      // Replace with your actual Klavis API key
  ```
</CodeGroup>

## Basic Setup

<CodeGroup>
  ```python Python theme={null}
  import os
  import json
  from together import Together
  from klavis import Klavis
  from klavis.types import McpServerName, ToolFormat

  # Initialize clients
  together_client = Together(api_key=os.getenv("TOGETHER_API_KEY"))
  klavis_client = Klavis(api_key=os.getenv("KLAVIS_API_KEY"))
  ```

  ```typescript TypeScript theme={null}
  import Together from 'together-ai';
  import { KlavisClient, Klavis } from 'klavis';

  // Initialize clients
  const togetherClient = new Together({ apiKey: process.env.TOGETHER_API_KEY });
  const klavisClient = new KlavisClient({ apiKey: process.env.KLAVIS_API_KEY });
  ```
</CodeGroup>

## AI Agent with MCP Integration

Now we'll create an intelligent agent that uses Together AI's powerful LLMs with Klavis MCP servers. This agent will:

1. **Discover Tools**: Automatically find available tools from MCP servers
2. **Function Calling**: Use Together AI's function calling capabilities
3. **Tool Execution**: Execute tools through Klavis API
4. **Smart Responses**: Generate intelligent responses based on tool results

<CodeGroup>
  ```python Python theme={null}
  class Agent:
      def __init__(self, together_client, klavis_client, mcp_server_url, model="meta-llama/Llama-3.3-70B-Instruct-Turbo"):
          self.together = together_client
          self.klavis = klavis_client
          self.mcp_server_url = mcp_server_url
          self.model = model
          print(f"🤖 Agent initialized with Together AI model: {self.model}")
      
      def process_request(self, user_message):
          # 1. Get available tools
          mcp_tools = self.klavis.mcp_server.list_tools(
              server_url=self.mcp_server_url,
      
              format=ToolFormat.OPENAI
          )
          
          # 2. Call LLM with tools
          messages = [
              {"role": "system", "content": "You are a helpful AI assistant with access to various tools."},
              {"role": "user", "content": user_message}
          ]
          
          response = self.together.chat.completions.create(
              model=self.model,
              messages=messages,
              tools=mcp_tools.tools
          )
          
          assistant_message = response.choices[0].message
          messages.append(assistant_message)
          
          # 3. If LLM wants to use tools
          if assistant_message.tool_calls:
              
              # Execute each tool call
              for tool_call in assistant_message.tool_calls:
                  tool_name = tool_call.function.name
                  tool_args = json.loads(tool_call.function.arguments)
                  
                  print(f"🛠️ Calling tool: {tool_name} with args: {tool_args}")
                  # Call tool via Klavis SDK
                  tool_result = self.klavis.mcp_server.call_tools(
                      server_url=self.mcp_server_url,
                      tool_name=tool_name,
                      tool_args=tool_args
                  )
                  
                  messages.append({
                      "role": "tool",
                      "tool_call_id": tool_call.id,
                      "content": str(tool_result)
                  })
              
              # 4. Get final response from LLM
              final_response = self.together.chat.completions.create(
                  model=self.model,
                  messages=messages
              )
              return final_response.choices[0].message.content
          
          # If no tools needed, return the assistant message directly
          return assistant_message.content
  ```

  ```typescript TypeScript theme={null}
  class Agent {
      private together: Together;
      private klavis: KlavisClient;
      private mcpServerUrl: string;
      private model: string;

      constructor(togetherClient: Together, klavisClient: KlavisClient, mcpServerUrl: string, model: string = "meta-llama/Llama-3.3-70B-Instruct-Turbo") {
          this.together = togetherClient;
          this.klavis = klavisClient;
          this.mcpServerUrl = mcpServerUrl;
          this.model = model;
          console.log(`🤖 Agent initialized with Together AI model: ${this.model}`);
      }

      async processRequest(userMessage: string) {
          // 1. Get available tools
          const mcpTools = await this.klavis.mcpServer.listTools({
              serverUrl: this.mcpServerUrl,
              format: Klavis.ToolFormat.Openai
          });

          // 2. Call LLM with tools
          const messages = [
              { role: "system", content: "You are a helpful AI assistant with access to various tools." },
              { role: "user", content: userMessage }
          ];

          const response = await this.together.chat.completions.create({
              model: this.model,
              messages: messages,
              tools: mcpTools.tools
          });

          const assistantMessage = response.choices[0].message;
          messages.push(assistantMessage);

          // 3. If LLM wants to use tools
          if (assistantMessage.tool_calls) {
              // Execute each tool call
              for (const toolCall of assistantMessage.tool_calls) {
                  const toolName = toolCall.function.name;
                  const toolArgs = JSON.parse(toolCall.function.arguments);

                  console.log(`🛠️ Calling tool: ${toolName} with args:`, toolArgs);
                  // Call tool via Klavis SDK
                  const toolResult = await this.klavis.mcpServer.callTools({
                      serverUrl: this.mcpServerUrl,
                      toolName: toolName,
                      toolArgs: toolArgs
                  });

                  messages.push({
                      role: "tool",
                      tool_call_id: toolCall.id,
                      content: JSON.stringify(toolResult)
                  });
              }

              // 4. Get final response from LLM
              const finalResponse = await this.together.chat.completions.create({
                  model: this.model,
                  messages: messages
              });
              return finalResponse.choices[0].message.content;
          }

          // If no tools needed, return the assistant message directly
          return assistantMessage.content;
      }
  }
  ```
</CodeGroup>

## Use Case Examples

### Example 1: Summarize YouTube Video

<Steps>
  <Step title="Initialize Clients">
    Set up Together AI and Klavis API clients
  </Step>

  <Step title="Create MCP Instance">
    Create a YouTube MCP server instance
  </Step>

  <Step title="Process Request">
    Use the agent to analyze and summarize a YouTube video
  </Step>
</Steps>

<CodeGroup>
  ```python Python theme={null}
  # Example YouTube video URL - replace with any video you'd like to analyze
  YOUTUBE_VIDEO_URL = "https://www.youtube.com/watch?v=TG6QOa2JJJQ"

  # 1. Create YouTube MCP server instance
  youtube_mcp_instance = klavis_client.mcp_server.create_server_instance(
      server_name=McpServerName.YOUTUBE,
      user_id="1234"
  )

  # 2. Create an agent with YouTube MCP server
  agent = Agent(
      together_client=together_client, 
      klavis_client=klavis_client, 
      mcp_server_url=youtube_mcp_instance.server_url,
      model="meta-llama/Llama-3.3-70B-Instruct-Turbo"
  )

  # 3. Process the request
  response = agent.process_request(
      f"Please analyze this YouTube video and provide a comprehensive summary with timestamps: {YOUTUBE_VIDEO_URL}"
  )

  print(response)
  ```

  ```typescript TypeScript theme={null}
  // Example YouTube video URL - replace with any video you'd like to analyze
  const YOUTUBE_VIDEO_URL = "https://www.youtube.com/watch?v=TG6QOa2JJJQ";

  // 1. Create YouTube MCP server instance
  const youtubeMcpInstance = await klavisClient.mcpServer.createServerInstance({
      serverName: Klavis.McpServerName.Youtube,
      userId: "1234"
  });

  // 2. Create an agent with YouTube MCP server
  const agent = new Agent(
      togetherClient, 
      klavisClient, 
      youtubeMcpInstance.serverUrl,
      "meta-llama/Llama-3.3-70B-Instruct-Turbo"
  );

  // 3. Process the request
  const response = await agent.processRequest(
      `Please analyze this YouTube video and provide a comprehensive summary with timestamps: ${YOUTUBE_VIDEO_URL}`
  );

  console.log(response);
  ```
</CodeGroup>

### Example 2: Send Email via Gmail

<Note>
  Gmail integration requires OAuth authentication, so you'll need to authorize the application in your browser.
</Note>

<CodeGroup>
  ```python Python theme={null}
  import webbrowser

  # Create Gmail MCP server instance
  gmail_mcp_instance = klavis_client.mcp_server.create_server_instance(
      server_name=McpServerName.GMAIL,
      user_id="1234"
  )

  # Redirect to Gmail OAuth page for authorization
  webbrowser.open(gmail_mcp_instance.oauth_url)
  print(f"🔐 Opening OAuth authorization for Gmail")
  print(f"If you are not redirected automatically, please open this URL: {gmail_mcp_instance.oauth_url}")

  # Email configuration
  EMAIL_RECIPIENT = "recipient@example.com"  # Replace with the recipient's email
  EMAIL_SUBJECT = "Greetings from Together AI + Klavis Integration"
  EMAIL_BODY = "This is a test email sent using the Together AI and Klavis AI integration. The email was sent automatically by your AI agent!"

  # After OAuth authorization is complete, create the Gmail agent
  gmail_agent = Agent(
      together_client=together_client,
      klavis_client=klavis_client,
      mcp_server_url=gmail_mcp_instance.server_url,
      model="Qwen/Qwen2.5-72B-Instruct-Turbo"
  )

  # Send the email
  response = gmail_agent.process_request(
      f"Please send an email to {EMAIL_RECIPIENT} with the subject '{EMAIL_SUBJECT}' and the following body: '{EMAIL_BODY}'"
  )

  print(response)
  ```

  ```typescript TypeScript theme={null}
  // Create Gmail MCP server instance
  const gmailMcpInstance = await klavisClient.mcpServer.createServerInstance({
      serverName: Klavis.McpServerName.Gmail,
      userId: "1234"
  });

  // Redirect to Gmail OAuth page for authorization
  console.log("🔐 Opening OAuth authorization for Gmail");
  console.log(`If you are not redirected automatically, please open this URL: ${gmailMcpInstance.oauthUrl}`);
  // In a web environment, you might redirect the user
  window.open(gmailMcpInstance.oauthUrl);

  // Email configuration
  const EMAIL_RECIPIENT = "recipient@example.com";  // Replace with the recipient's email
  const EMAIL_SUBJECT = "Greetings from Together AI + Klavis Integration";
  const EMAIL_BODY = "This is a test email sent using the Together AI and Klavis AI integration. The email was sent automatically by your AI agent!";

  // After OAuth authorization is complete, create the Gmail agent
  const gmailAgent = new Agent(
      togetherClient,
      klavisClient,
      gmailMcpInstance.serverUrl,
      "Qwen/Qwen2.5-72B-Instruct-Turbo"
  );

  // Send the email
  const response = await gmailAgent.processRequest(
      `Please send an email to ${EMAIL_RECIPIENT} with the subject '${EMAIL_SUBJECT}' and the following body: '${EMAIL_BODY}'`
  );

  console.log(response);
  ```
</CodeGroup>

## Next Steps

<CardGroup cols={2}>
  <Card title="Explore More MCP Servers" icon="server">
    Try other available servers like Slack, Notion, CRM etc.
  </Card>

  <Card title="Experiment with Different Models" icon="code">
    Test various Together AI models for different use cases.
  </Card>

  <Card title="Build Multi-Server Workflows" icon="laptop">
    Create sophisticated agents that combine multiple services
  </Card>

  <Card title="Production Deployment" icon="rocket">
    Scale these patterns for production applications
  </Card>
</CardGroup>

## Useful Resources

* [Together AI Documentation](https://docs.together.ai/)
* [Klavis AI Documentation](https://www.klavis.ai/docs/)
* [MCP Protocol Specification](https://modelcontextprotocol.io/)
* [Together AI Models](https://docs.together.ai/docs/inference-models)
* [Klavis MCP Servers](/mcp-server)

**Happy building with Together AI and Klavis!** 🚀