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

# Agno

> Build autonomous multimodal agents, multi-agent systems, and tool-driven workflows with Agno.

[Agno](https://www.agno.com/?utm_source=neosantara-docs\&utm_medium=referral) is an open-source framework for building autonomous agents, multi-agent systems, and multimodal workflows. Neosantara is available as a native model provider via `agno.models.neosantara.Neosantara`, enabling tool calling, persistent memory, and multimodal reasoning powered natively by Indonesia's AI Gateway.

## Installation

Install Agno and required tools:

```bash theme={"theme":{"light":"ayu-dark","dark":"catppuccin-latte"}}
pip install -U agno ddgs
```

## Authentication

Set your Neosantara API key in your environment:

```bash theme={"theme":{"light":"ayu-dark","dark":"catppuccin-latte"}}
export NEOSANTARA_API_KEY="nsk_..."
```

The native Agno provider automatically targets `https://api.neosantara.xyz/v1`.

## First Native Agent

<CodeGroup>
  ```python Python (Basic Agent) icon="python" theme={"theme":{"light":"ayu-dark","dark":"catppuccin-latte"}}
  from agno.agent import Agent
  from agno.models.neosantara import Neosantara

  agent = Agent(
      model=Neosantara(id="claude-opus-4-6"),
      markdown=True
  )

  agent.print_response("Explain the role of an AI gateway in 2 sentences.")
  ```

  ```python Python (Streaming Output) icon="python" theme={"theme":{"light":"ayu-dark","dark":"catppuccin-latte"}}
  from agno.agent import Agent
  from agno.models.neosantara import Neosantara

  agent = Agent(
      model=Neosantara(id="gemini-3.8-flash"),
      markdown=True
  )

  agent.print_response(
      "Create an incident response checklist for distributed systems.",
      stream=True
  )
  ```
</CodeGroup>

## Agent with Tool Calling (Search)

Equip native Neosantara models with external web search tools:

```python theme={"theme":{"light":"ayu-dark","dark":"catppuccin-latte"}}
from agno.agent import Agent
from agno.models.neosantara import Neosantara
from agno.tools.duckduckgo import DuckDuckGoTools

agent = Agent(
    model=Neosantara(id="gemini-3.8-flash"),
    tools=[DuckDuckGoTools()],
    instructions="Always provide verified source citations.",
    markdown=True
)

agent.print_response("What is the Southeast Asian AI startup market valuation in 2026?")
```

## Agno Integration Advantages

| Capability                      | Generic OpenAIChat Wrapper                  | Native Neosantara Provider                       |
| :------------------------------ | :------------------------------------------ | :----------------------------------------------- |
| **Module Import**               | `from agno.models.openai import OpenAIChat` | `from agno.models.neosantara import Neosantara`  |
| **Tool Calling & Agent Memory** | Dependent on generic wrapper translation    | Natively bound to Agno agent execution loops     |
| **Model Support**               | Requires manual parameter overrides         | Supports the complete Neosantara model catalog   |
| **Streaming & Markdown**        | Manual response extraction                  | Built-in markdown formatting and token streaming |


## Related topics

- [Integrations Overview](/en/integrations/overview.md)
- [CrewAI](/en/integrations/crewai.md)
- [MCP & Agent Architecture](/en/agents/overview.md)
- [Model Catalog](/en/gateway/models.md)