dartantic_gemma 0.3.0

Dartantic provider plugin enabling local LLM inference using Google's Gemma models via Flutter

Changelog

All notable changes to this project will be documented in this file.

The format is based on Keep a Changelog, and this project adheres to Semantic Versioning.

0.3.0 - 2026-05-26

• Updated flutter_gemma dependency to ^0.16.0 for latest native FFI improvements
• NPU dispatch now available on Linux/macOS alongside Windows
• Desktop storage updated to use Application Support instead of Documents
• Support for qdrant-edge vector store (replaces sqlite + local_hnsw)

0.2.4 - 2026-04-30

• Added handling of Gemma 4 tool calling format parsing to properly register function executions
• Update minimum flutter_gemma version to ^0.14.0

0.2.3 - 2026-04-13

Fixed

• Fixed tool call ID mapping in GemmaChatModel.sendStream by adding a unique toolCallIdCounter to ensure unique IDs for each tool call, preventing orchestration errors in multi-iteration agent loops.

0.2.2 - 2026-04-13

Fixed

• Improved session management in GemmaChatModel by creating a fresh chat instance for each sendStream call, preventing "Session is closed" errors in multi-iteration agent loops.
• Enhanced tool result formatting for Gemma 2 using <start_of_role>tool<end_of_role> tags.

0.2.1 - 2026-04-13

Fixed

• Fixed model being closed prematurely in GemmaChatModel.dispose, which caused errors in multi-iteration agent loops.

0.2.0 - 2026-04-13

Added

• Support for selecting preferred backend (CPU/GPU) in FlutterGemmaProvider and GemmaChatModelOptions.
• Export PreferredBackend, ModelType, and ModelFileType from flutter_gemma.

0.1.0 - 2026-04-13

Added

• Initial release of dartantic_gemma package
• Dartantic provider plugin for Flutter Gemma local LLM inference
• Integration with dartantic_interface ^4.0.0
• Support for Flutter Gemma ^0.13.2 and genai_primitives ^0.2.0