RAGAS

This release focuses heavily on migrating core metrics to the new collections API structure and introduces caching support for metrics and embeddings. Several bug fixes address issues related to instructor modes, type validation, and Claude workflow tokens.

This release focuses heavily on migrating core metrics (ToolCallAccuracy, ToolCallF1, TopicAdherence, AgentGoalAccuracy, Rubrics) to utilize the collections API for better structure. It also introduces a breaking change by renaming `embed_text` to `aembed_text` in AnswerRelevancy.

This release introduces major architectural updates, migrating numerous metrics to a modular BasePrompt system and enhancing LLM provider support via instructor.from_provider and dual adapter capabilities. It also includes several bug fixes related to LangChain integration and LLM detection.

This release focuses heavily on migrating core metrics to a new structure, removing deprecated metrics like 'aspect critic', and introducing new features like synthetic data traceability metadata. Several documentation fixes and minor bug fixes related to OpenAI models were also implemented.

This release focuses heavily on internal refactoring, migrating core functionalities like semantic similarity and simple criteria to collections, and merging LLM factory methods. Several bugs related to async handling and specific synthesizers were also fixed.

This release focuses on migrating several core metrics (BLEU, string metrics, answer similarity) to collections, improving robustness in query distribution, and adding new configuration options for LLM wrappers. Internal code quality and documentation were also enhanced.

This release introduces several new features, including CHRF score support, enhanced input flexibility for metrics, and OCI Gen AI integration. Numerous bug fixes address issues related to asyncio, metric calculations, and dependency compatibility.

This release focuses on improving core functionality, including better async execution and knowledge graph optimization, alongside several bug fixes and documentation updates.

No release notes provided.

This release focuses on improving asynchronous operations, optimizing knowledge graph handling for large datasets, and fixing a TypeError in metric calculations. It also introduces telemetry collection.

This release focuses on performance improvements, documentation updates, and minor bug fixes, including optimizing cluster finding and fixing batching issues with LangChain.

This release focuses heavily on internal restructuring, moving modules like `tracing`, `prompts`, `dataset`, and experimental features into the main package structure while retiring the `ragas.experimental` namespace. Numerous bug fixes address CI, LLM compatibility (especially OpenAI O1 series), and metric stability.

This release focuses heavily on internal restructuring, migrating modules like `tracing`, `prompts`, `dataset`, and experimental metrics out of experimental namespaces and into the main package structure. It also includes numerous bug fixes, performance optimizations (like 50% speedup for factual correctness), and improved LLM compatibility.

This release moves key features like `experiment` and the CLI from experimental to the main package, adds prompt saving/loading capabilities, and removes the simulation feature.

No release notes provided.

This release (v0.3.2-rc2) primarily addresses fixes related to pypi requirements and image absolute paths.

This release moves key features like `experiment` and the CLI from experimental to the main package, removes simulation functionality, and adds support for Python 3.13.

This release introduces a new Google Drive backend for dataset storage and includes several documentation and example improvements, alongside minor configuration fixes.

This release introduces major features like LlamaIndex agentic integration, a new CLI, and security enhancements including a fix for CVE-2025-45691. It also includes significant internal refactoring, notably the removal of the Project structure.

No release notes provided.

No release notes provided.

This release introduces new integrations with AWS Bedrock, LlamaStack, and Griptape, alongside enhancements to validation logic and documentation updates. A key documentation change involves renaming AWS Bedrock references to Amazon Bedrock.

This release introduces new features like HTTP request-response logging and multi-turn conversation evaluation, alongside numerous bug fixes across various metrics and synthesizers. It also includes documentation updates and new integrations.

This release focuses on bug fixes, prompt improvements, and enhancements to integrations like langgraph, alongside removing an unnecessary argument from ToolCallAccuracy initialization.

This release introduces Bedrock token parser support and an optional parameter for the BLEU score, alongside several bug fixes for TP/FP calculations and the output parser.

This release introduces new features like Swarm integration and the ability to specify an experiment name during evaluation. It also includes several bug fixes related to metrics and dependency management, alongside numerous documentation updates.