fal.ai

The [fal.ai] inference API is integrated into promptfoo as a provider using the fal-js client, providing a native experience for using fal.ai models in your evaluations.

To run a model, specify the model type and model name: fal:<model_type>:<model_name>.

Supported models include

• fal:image:fal-ai/fast-sdxl
• fal:image:fal-ai/flux/dev
• fal:image:fal-ai/flux/schnell

Browse our model gallery for a list of available models.

Setup

To use fal, you need to set up your API key:

1. Install the fal client as a dependency:

   npm install -g @fal-ai/serverless-client
2. Create am API key in the fal dashboard.
3. Set the FAL_KEY environment variable:

   export FAL_KEY=your_api_key_here

Alternatively, you can specify the apiKey in the provider configuration (see below).

Configuration

Configure the fal image provider in your promptfoo configuration file. Here's an example using fal-ai/flux/schnell:

providers:
  - id: fal:image:fal-ai/flux/schnell
    config:
      image_size:
        width: 1024
        height: 1024
      num_inference_steps: 8
      seed: 6252023

Key configuration options:

Parameter	Description
apiKey	The API key for authentication with fal.

Common image generation parameters:

Parameter	Type	Description
image_size.width	number	The width of the generated image.
image_size.height	number	The height of the generated image.
num_inference_steps	number	The number of inference steps to run.
seed	number	Sets a seed for reproducible results.

warning

Not every model supports every parameter. Check the API docs of each model for more details.