Fal

FalTools enable an Agent to perform media generation tasks.

Prerequisites

The following example requires the fal_client library and an API key which can be obtained from Fal.

pip install -U fal_client

export FAL_KEY=***

Example

The following agent will use FAL to generate any video requested by the user.

cookbook/tools/fal_tools.py

from agno.agent import Agent
from agno.models.openai import OpenAIChat
from agno.tools.fal import FalTools

fal_agent = Agent(
    name="Fal Video Generator Agent",
    model=OpenAIChat(id="gpt-5-mini"),
    tools=[FalTools("fal-ai/hunyuan-video")],
    description="You are an AI agent that can generate videos using the Fal API.",
    instructions=[
        "When the user asks you to create a video, use the `generate_media` tool to create the video.",
        "Return the URL as raw to the user.",
        "Don't convert video URL to markdown or anything else.",
    ],
    markdown=True,
    debug_mode=True,
    )

fal_agent.print_response("Generate video of balloon in the ocean")

Toolkit Params

Parameter	Type	Default	Description
api_key	str	None	API key for authentication purposes.
model	str	None	The model to use for the media generation.
enable_generate_media	bool	True	Enable the generate_media functionality.
enable_image_to_image	bool	True	Enable the image_to_image functionality.
all	bool	False	Enable all functionality.

Toolkit Functions

Function	Description
generate_media	Generate either images or videos depending on the user prompt.
image_to_image	Transform an input image based on a text prompt.

Developer Resources

• View Tools
• View Cookbook