saptiva-agents 0.1.4

Saptiva-Agents es un potente framework para construir aplicaciones de IA multiagente que pueden actuar de forma autónoma o colaborar con humanos.

🤖 Saptiva Agents

Saptiva-Agents is a powerful framework for building multi-agent AI applications that can act autonomously or collaborate with humans.

---

⚙️ Installation

Saptiva-Agents requires Python 3.10 or later. To install from PyPI:

pip install -U saptiva-agents

---

🚀 Quickstart

👋 Hello World

Create an assistant agent using llama3.3:70b with Saptiva-Agents:

import asyncio

from saptiva_agents import LLAMA_MODEL
from saptiva_agents.base import SaptivaAIChatCompletionClient
from saptiva_agents.agents import AssistantAgent

async def main() -> None:
    model_client = SaptivaAIChatCompletionClient(
        model=LLAMA_MODEL, 
        api_key="YOUR_SAPTIVA_API_KEY"
    )
    agent = AssistantAgent("assistant", model_client=model_client)
    print(await agent.run(task="Say 'Hello World!'"))
    await model_client.close()

asyncio.run(main())

python hello_world.py

---

🌐 Web Browsing Agent Team

Create a browser-based agent team using Playwright:

# pip install saptiva-agents
# playwright install --with-deps chromium

import asyncio

from saptiva_agents import QWEN_MODEL
from saptiva_agents.agents import UserProxyAgent
from saptiva_agents.base import SaptivaAIChatCompletionClient
from saptiva_agents.conditions import TextMentionTermination
from saptiva_agents.teams import RoundRobinGroupChat
from saptiva_agents.web_surfer import MultimodalWebSurfer
from saptiva_agents.ui import Console

async def main() -> None:
    model_client = SaptivaAIChatCompletionClient(
        model=QWEN_MODEL, 
        api_key="YOUR_SAPTIVA_API_KEY"
    )
    web_surfer = MultimodalWebSurfer(
        "web_surfer", 
        model_client, 
        headless=False, 
        animate_actions=True, 
        start_page="https://www.google.com"
    )
    user_proxy = UserProxyAgent("user_proxy")
    termination = TextMentionTermination("exit", sources=["user_proxy"])
    team = RoundRobinGroupChat([web_surfer, user_proxy], termination_condition=termination)

    try:
        await Console(team.run_stream(task="Navega a saptiva.com y consigue información sobre Saptiva AI."))
    finally:
        await web_surfer.close()
        await model_client.close()

asyncio.run(main())

python web_surfer.py

---

📚 Available Models

Saptiva-Agents supports a variety of models for both text and multimodal tasks.

🧠 Text Models

Model	Description
llama3.3:70b	State-of-the-art model based on Meta LLaMA 3, ideal for complex natural language tasks.
qwen2.5:72b-instruct	Advanced instruction-tuned model from Alibaba DAMO, designed to follow user prompts precisely.

🖼️ Multimodal Models

Model	Description
gemma3:27b	Multimodal model capable of reasoning over visual and textual inputs. Ideal for interactive tasks and web exploration.

🔗 View the full list of available Saptiva models

---

📄 License

This project is licensed under the MIT License.

---

🌐 Links

• 🔗 Official Docs