Validates and coerces unstructured LLM responses directly into rich, schema-validated Ruby objects with automatic self-correction loops.

A mode is the configuration of one turn: a RubyLLM agent with a routing description. Declare the modes, a fallback, and a classifier; before each answer the router returns a Route that says which mode takes the turn, why, and what the classifier actually said.

Drop-in Mongoid replacement for ruby_llm's ActiveRecord integration. Provides acts_as_chat, acts_as_message, acts_as_tool_call, and acts_as_model macros backed by MongoDB via Mongoid.

A Rails engine providing comprehensive observability for LLM-powered applications. Features include session tracking, trace analysis, prompt management, cost monitoring, and optional chat/agent testing UI (with RubyLLM integration).

Adds Pollinations AI support to RubyLLM — chat, image/video generation, TTS/music, and transcription via the Pollinations API. Installs as a standalone gem without modifying RubyLLM core.

Adds Apfel provider support to RubyLLM. Apfel is a macOS-only local server that exposes Apple's on-device Apple Intelligence model over an OpenAI-compatible API, letting RubyLLM talk to it like any other provider.

Adds LM Studio provider support to RubyLLM via LM Studio's OpenAI-compatible local server, with model listings enriched from LM Studio's native REST API.

Typed judgments and reranking with the TypeSafe Jev model through RubyLLM.

A RubyLLM plugin that enables local LLM execution using the Red Candle gem. Run quantized GGUF models directly in Ruby without external API calls. Supports streaming, structured output, and multiple model architectures including Gemma, Llama, Qwen, Mistral, and Phi.

Cache RubyLLM responses based on semantic similarity, not exact string matching. Reduces costs and latency by returning cached responses for semantically equivalent queries.

Route messages to specialized RubyLLM chat agents based on semantic similarity using embeddings and kNN lookup

Provides an adapter to integrate RubyLLM with Sequel ORM.

A RubyLLM extension that groups named coworkers and exposes delegation tools so a model can hand tasks and questions to teammates. Built on the RubyLLM public API: Tool, Agent, Message, and Attachment.

Provides a RubyLLM::Provider that allows you to stub responses for testing purposes. You can stub individual responses or a sequence of responses, and you can also temporarily stub responses within a block.

Intuitive one-liner utility methods for common LLM tasks like text summarization, translation, data extraction, classification, grammar correction, sentiment analysis, key point extraction, text rewriting, and question answering.

ruby_llm-toolbox bundles filesystem, shell, web, git, and data tools as ready-to-use RubyLLM::Tool subclasses. One require loads everything; read-only tools are on by default while mutating/exec tools are gated behind explicit config. Tool output is token-budgeted via ruby_llm-tokenizer.

A RubyLLM plugin that integrates Top Secret to automatically filter sensitive information from LLM interactions.

Adds a :typesafe provider to RubyLLM 2 that evaluates text or structured state against TypeSafe's typed Choice, Noul, and Score questions through Chat#with_schema. Structured output only: answers are calibrated probabilities, not generated text.

Automatically report AI agent health, response times, and model info to your ups.dev status page.

A RubyLLM extension for Voyage embeddings, reranking, contextualized chunks, and files.