gemini.R (version 0.16.0)

gemini_audio: Analyze audio using Gemini

Description

This function sends audio to the Gemini API and returns a text description.

Usage

gemini_audio(
  audio = NULL,
  prompt = "Describe this audio",
  model = "2.0-flash",
  temperature = 1,
  maxOutputTokens = 8192,
  topK = 40,
  topP = 0.95,
  seed = 1234
)

Value

A character vector containing the Gemini API's response.

Arguments

audio

Path to the audio file (default: uses a sample file). Must be an MP3.

prompt

A string describing what to do with the audio.

model

The model to use. Options are "2.0-flash", "2.0-flash-lite", "2.5-pro-exp-03-25". Default is '2.0-flash' see https://ai.google.dev/gemini-api/docs/models/gemini

temperature

The temperature to use. Default is 1 value should be between 0 and 2 see https://ai.google.dev/gemini-api/docs/models/generative-models#model-parameters

maxOutputTokens

The maximum number of tokens to generate. Default is 8192 and 100 tokens correspond to roughly 60-80 words.

topK

The top-k value to use. Default is 40 value should be between 0 and 100 see https://ai.google.dev/gemini-api/docs/models/generative-models#model-parameters

topP

The top-p value to use. Default is 0.95 value should be between 0 and 1 see https://ai.google.dev/gemini-api/docs/models/generative-models#model-parameters

seed

The seed to use. Default is 1234 value should be integer see https://ai.google.dev/gemini-api/docs/models/generative-models#model-parameters

Details

The API key is now sent via the HTTP header x-goog-api-key instead of as a URL query parameter.

Examples

if (FALSE) {
library(gemini.R)
setAPI("YOUR_API_KEY")
gemini_audio(audio = "YOUR_AUDIO_FILE")
}