airi Releases

This release adds internationalization support for Spanish and introduces new AI providers including LM Studio and a generic OpenAI-compatible option. It also features a new unified model selector for managing different model types.

Aug 24, 2025

This update fixes the auto eye blink functionality and introduces support for Index-TTS. A keyboard shortcut for window resizing has been changed, and users can now switch between hemisphere lighting and skybox environments.

Aug 21, 2025

This update introduces an immersive UI experience, adds new compiled JavaScript to the source code, and includes a GPT-OSS demo.

Aug 12, 2025

A minor maintenance release that addresses a specific bug preventing the resource island from functioning correctly.

Aug 5, 2025

This major version introduces a complete build system overhaul using Turborepo and ROLLDOWN, along with the migration to VitePress for documentation. It loads Whisper and VAD models for voice capabilities and adds comprehensive i18n support.

Aug 4, 2025

This release enables users to programmatically close the connection using Python code, improving control over the application lifecycle.

Jul 7, 2025

This release focuses on bug fixes for API key detection and IME input, alongside new features for character card creation, Player2 API support, and VRM lip-sync.

Jun 17, 2025

A major update focusing on UI improvements, documentation, and stage functionality including moving the Tamagotchi chat to a standalone window and adding Live2D mouse tracking.

Jun 8, 2025

Introduced prompt engineering playground, basic MCP support, and new UI components such as Select and PageHeader to enhance the user interface.

May 10, 2025

Continuation of v0.5.0 features, specifically implementing basic MCP support, refactoring the Tamagotchi to use a top-level cargo workspace, and adding the Select component.

May 1, 2025

Focuses on UI components and prompt engineering features, including the prompt engineering playground, access to velin, and the Select component.

Apr 27, 2025

Introduces new trainers for control conditioning and adds support for FLUX.1-dev and Wan2.1 I2V models, alongside enhanced attention provider support.

Apr 25, 2025

A major UI overhaul featuring a motion mapping editor, theme customization, and Twitter integration, along with file system support via Drizzle and DuckDB-WASM.

Apr 7, 2025

Initial stable release introducing support for CogView4 and Wan 2.1 models, along with improved dataset loading and Accelerate compatibility.

Mar 21, 2025

This release contains no significant changes.

Feb 28, 2025

Added transformers.js package integration for AI capabilities and improved window management by allowing Tamagotchi windows to be resized and saved.

Feb 27, 2025

Introduced a shortcuts editor, system tray support, and file system support using Drizzle and DuckDB-WASM.

Feb 26, 2025

Initial public release of the library to support training diffusion models (CogVideoX, LTX Video, Hunyuan Video) with LoRA and full-rank finetuning.

Feb 24, 2025

A minor maintenance release fixing the tamagotchi distribution path in the CI configuration.

Feb 20, 2025

This release adds animated idle eye movements to Live2D models and implements internationalization (i18n) for the stage tamagotchi feature, alongside a fix for pnpm filtering on Windows.

Feb 20, 2025