Ruby client for Gemini generateContent. Includes a CLI and a PR review app.

gem install nuntius-rb

With Bundler:

gem 'nuntius-rb', require: 'nuntius'

The package is published as nuntius-rb. The runtime entrypoint is nuntius.

Set your API key in .env:

GEMINI_API_KEY=your_api_key

require 'nuntius'
Nuntius.load_env

client = Nuntius::Client.new
puts client.generate_text('Write a haiku about Ruby')

Use a different model when needed:

fast_client = Nuntius::Client.new(model: :flash)
puts fast_client.generate_text('Explain Ruby in one sentence')

Short aliases: :flash uses gemini-3.8-flash (default), :flash_fallback uses gemini-3.7-flash, and :flash_lite uses gemini-3.1-flash-lite. Legacy :pro_2_0 maps to gemini-2.0-flash. Pro models (gemini-pro-latest, gemini-3.1-pro-preview, gemini-2.5-pro) are retired for Free-Tier, as they resolve to the 3.1-pro backend and hit quota 0. Use billing to enable.

The gem does not wrap embeddings, Imagen, or Veo APIs.

Nuntius supports text generation, chat, image input for generateContent, model aliases, safety settings, API key masking, retries, and a local CLI.

Client validation and API failures raise Nuntius::Error with a readable message. HTTP 429 responses are retried automatically up to three times with exponential backoff.

begin
  response = client.generate_text('Hello')
  puts response
rescue Nuntius::Error => err
  warn "Generation failed: #{err.message}"
end

Common failures include:

• Missing or invalid GEMINI_API_KEY
• Empty prompts
• Prompts over the configured maximum length
• Gemini API errors returned by the service
• Network errors raised by HTTParty

Rate-limit responses (429) are retried up to three times with waits of 5, 10, and 20 seconds.

Requests use a 30 second HTTParty timeout.

require 'nuntius'

Nuntius::Client.logger.level = Logger::INFO
client = Nuntius::Client.new

Ruby 3.3 or later. Linux and macOS are tested.

GEMINI_API_KEY=your_api_key_here

./bin/gemini test
./bin/gemini generate "Your prompt"
./bin/gemini chat

bundle exec rake test          # Run tests
bundle exec rake docs          # Build API docs
gem build nuntius-rb.gemspec

Nuntius Review is the PR review app in this repo. It defaults to gemini-3.8-flash (fallback gemini-3.7-flash for large diffs). Override with NUNTIUS_GEMINI_MODEL. Retrieval context is off unless enabled in nuntius/config.yaml. Only verified generateContent Flash models are listed as supported. See API GET /v1beta/models for existence, not quota.

For setup details, see nuntius/Nuntius.md.

client = Nuntius::Client.new
puts client.generate_text('Write a haiku about Ruby')

image_data = Base64.strict_encode64(File.binread('path/to/image.jpg'))
puts client.generate_image_text(image_data, 'Describe this image')

messages = [
  { role: 'user', content: 'Hello!' },
  { role: 'model', content: 'Hi there!' },
  { role: 'user', content: 'Tell me about Ruby.' }
]
puts client.chat(messages, system_instruction: 'Be concise.')

Fork the repo and open a pull request.

MIT → see LICENSE.

Nuntius Review