1.0.1

New Features:

• General code improvements and library updates
• Add mail subject to the context
• Added new error checking functionality
• Added handling of LLM responses in markdown
• Add support for GPT-5.4

1.0

First stable version 🎉

New Features:

• General code improvements and library updates
• Added the ability to configure the model on Mistral
• Added PII masking option
• Modified default temperature value 0.25 for more deterministic responses
• Modified quality scripts to lint (semantic consistency)
• Added quick actions in LLM textual results
• Added option to run unit tests for a single provider
• OpenAI: Migrated to REST messages service, replacing chat/completions

Fix:

• Fix "CORS Error" in OpenAI API call when the organization is set
• Wrong URL for the scripting.messageDisplay API

0.94

New Features:

• General code improvements and library updates
• Anthropic: Update to Claude 4.5 Opus and deprecation of legacy models
• Google: Update to Gemini 3 and deprecation of legacy models
• OpenAI: Update to GPT-5.2 and deprecation of legacy models
• xAI: Update to Grok 4.1 and deprecation of legacy models
• Minor updates to documentation and unit tests

0.93

New Features:

• Added support for using custom prompts in both email viewing and composition windows;
• Updated xAI Grok model (Grok 4);
• Increased the minimum supported Thunderbird version to 128+, enabling access to new APIs;
• Replaced the sanitize-html library with the native convertToPlainText function from the messengerUtilities API, significantly reducing the add-on size;
• Performed internal code optimizations;
• Updated third-party libraries.

Fixes:

• Improved Mistral API error handling.

0.92

New Features:

• Added the ability to analyze the intent of emails being composed;
• Added functionality to remove any Markdown syntax from textual responses;
• Updated Claude models (Claude 4 Opus and Claude 4 Sonnet) and removed deprecated models (Claude 3 Opus and Claude 3 Sonnet);
• Updated Google Gemini models (stable 2.5 Pro and 2.5 Flash);
• Updated OpenAI models (GPT-4.1 and GPT-4.1 mini);
• Updated xAI Grok model (stable Grok 3, Grok 3 Fast, Grok 3 Mini and Grok 3 Mini Fast).

Fixes:

• Improved Mistral API error handling.

0.91

New Features:

• Possibility to configure a set of languages to manage the translation of an email;
• Added email explanation feature;
• Added support for Mistral AI;
• Update of Claude models with the final version of Claude 3.7 Sonnet;
• Update of Gemini models with the final version of 2.0 Flash, 2.0 Flash Lite and 2.5 Pro (Experimental);
• Update of Grok models with Grok 3 (Beta) and Grok 3 Fast (Beta);
• Redefinition of all prompts to improve the behavior of LLM agents;
• Updated libraries.

Fixes:

• Fixed incorrect behavior when selecting a tone of voice during reply suggestion.

0.90

New Features:

• Added functionality to receive improvement suggestions;
• Added support for DeepSeek;
• Added support for LM Studio;
• Added new voices for OpenAI Text-to-Speech (Ash, Coral, and Sage);
• Updated model references for Claude 3.5 Haiku and Grok xAI;
• Improved web component for API key management;
• Updated libraries.

Fixes:

• Fixed incorrect labels for Groq Cloud;
• Removed unused code.

0.89

New Features:

• Removal of the Google Gemini 1.0 Pro model, now deprecated and scheduled for removal on February 15, 2025;
• Added support for Grok by xAI;
• Added support for Gemini 1.5 Flash-8B model and Gemini 2.0 Flash model (Experimental);
• Library updates.

Fixes:

• Fix for temperature parameter handling in Google Gemini;
• Owl for Exchange add-on not detected.

0.88

New Features:

• Updated Claude models;
• Integration with Ollama (https://ollama.com);
• Integration with GroqCloud (https://groq.com);
• Library updates;
• Minor general code optimizations.

Fixes:

• Updated CSS styles to resolve certain cases where responses were not displaying;
• Fixed incorrect configuration message in the menu for “Suggest Reply”.

0.87

New features:

• Added support for Google Gemini LLM;
• Added temperature option;
• Added 9 options to manage tone of voice when rephrasing a portion of text or suggesting a reply to an email;
• General improvements in result handling;
• Improved reading of email text components;
• Library updates;
• Minor general code optimizations.