What does this prompt chain do?

It demonstrates using Google's Live API for audio generation with Gemini models and evaluates the speech output through promptfoo's standard evaluation flow.

What environment variable do I need to run this?

You need a `GOOGLE_API_KEY` environment variable, which you can obtain from the Google AI Studio website.

How do I set up and run this example?

Run `npx promptfoo@latest init --example google-live-audio`, then execute `promptfoo eval -c promptfooconfig.yaml` to generate audio outputs, and use `promptfoo view` to see the results.

When should I use this instead of text-only Gemini evaluation?

Use this when you specifically want to test Gemini's Live API audio-generation capability; it is not suitable if you only need text-only Gemini evaluation.

Prompt Chain

Generate Audio with Google Live API

Name: Google Live Audio
Availability: OnlineOnly
Author: Promptfoo

Promptfoo example evaluating audio generation through Google's Live API with Gemini models.

Copy chain

Works with google

Promptfoo

Own this? Claim it

Spark score

out of 100

Updated 22 days ago

Version 0.121.18

Models

gemini 2 0

Add to Favorites

Why it matters

Leverage Google's Live API and Gemini models to generate audio content. This prompt chain facilitates the creation of audio outputs through a structured process.

Outcomes

What it gets done

Generate audio using Google's Live API

Utilize Gemini models for audio creation

Process audio through a prompt chain

Install

Add it to your toolbox

Run in your project directory:

curl -fsSL https://spark.entire.vc/get/pfoo-google-live-audio | bash

Steps

Steps in the chain

Initialize google-live-audio example

Navigate to project directory

Run evaluation

View results

Overview

Google Live Audio

This promptfoo example evaluates audio generation through Google's Live API with Gemini models, using a minimal ready-to-run config. Use it when you need to test Gemini's Live API audio-generation capability with promptfoo; not applicable to text-only Gemini evaluation.

What it does

This promptfoo example demonstrates using Google's Live API for audio generation with Gemini models, evaluating speech output through promptfoo's standard eval flow.

When to use - and when NOT to

Use it when you want to test Gemini's Live API audio-generation capability with promptfoo. Do not use it if you need text-only Gemini evaluation - this example is scoped specifically to Live API audio generation.

Inputs and outputs

Requires a GOOGLE_API_KEY environment variable, obtainable from the Google AI Studio website. Running promptfoo eval -c promptfooconfig.yaml generates audio outputs from the configured test cases; results are viewable with promptfoo view.

Integrations

Runs via npx promptfoo@latest init --example google-live-audio, then promptfoo eval -c promptfooconfig.yaml and promptfoo view.

Who it's for

Developers evaluating Gemini's Live API for audio and speech generation who need a minimal working promptfoo config to start from.

Source README

google-live-audio (Google Live API Audio with Gemini)

This example demonstrates how to use promptfoo with Google's Live API for audio generation using Gemini models.

You can run this example with:

npx promptfoo@latest init --example google-live-audio
cd google-live-audio

Prerequisites

Google AI Studio API key set as the environment variable GOOGLE_API_KEY

You can obtain a Google AI Studio API key from the Google AI Studio website.

Running the Example

promptfoo eval -c promptfooconfig.yaml

View the results:

promptfoo view

For more information about the Google Live API, see the Google AI Speech Generation documentation.

FAQ

Common questions

Discussion