RubyLLM 1.0

RubyLLM is a Ruby native framework for building AI features in plain Ruby or Rails applications. It gives developers one consistent API for working with language models, tools and agents across 19 providers, so an application can move between hosted and local models without rewriting its code. The listing names version 1.0, while the site documents version 2.1.1 as the current stable release. The library covers the common jobs of an LLM application. A chat object keeps conversation history, and responses can be streamed to a terminal or a Rails view. Files such as images, recordings and PDFs can be passed to a model by local path or URL. Developers describe tools as Ruby classes and RubyLLM runs the tool calls and returns the results to the model. It can also connect to Model Context Protocol servers, and it supports agents, image generation, audio and video, structured output and usage and cost tracking. Providers include Anthropic, Ollama and Hetzner, and any OpenAI compatible endpoint can be connected. A provider gem generator scaffolds support for services that are not yet covered. The site also hosts a model registry listing more than 1,800 models across 17 providers, with capabilities, context limits and standard text token prices in US dollars per million tokens, available as JSON and refreshable from Ruby. Installation is through a gem with bundle add. The site publishes no separate pricing for the library itself, since model usage is billed by whichever provider is selected.