TalkingGaussian Releases
86 releases of Fictionarry/TalkingGaussian
- 0.77.0
This release adds support for controlling physical iOS devices alongside simulators, introduces a new `instruction` command for manual scenario reproduction, and upgrades the underlying Maestro engine to version 2.7.
Jul 23, 2026
- v3.6.0
This release focuses on stability and security with comprehensive hardening across the research pipeline. It introduces integration with LangChain's Deep Agents framework and adds Nebius Token Factory as a new LLM provider.
Jul 18, 2026
- v3.0.23
Fixed a critical bug where Vibe and Sona windows remained open on Windows after exiting, which could block relaunching. Added an advanced setting to automatically unload inactive transcription models to free up RAM and VRAM. Updated Sona to v0.3.5 with improved process monitoring.
Jul 15, 2026
- model-files-v1.0
Jul 13, 2026
- 0.76.1
Corrected a rendering issue in the scenario graph where dependency edges were incorrectly drawn for reusable calls. The visual representation now accurately reflects the execution order by drawing the dependency edge from the last reusable call node.
Jul 13, 2026
- 0.76.0
Introduced a new Scenario Graph feature to visualize the execution structure of projects, available both in the CLI and UI. Added built-in guides for AI coding agents to help them operate the tool end-to-end. Added scenario search functionality in the UI and upgraded the underlying toolchain.
Jul 13, 2026
- 0.75.0
Introduced Reusable Scenarios, allowing scenarios to be defined once and called with parameters from anywhere, similar to GitHub Actions workflows. Added direct support for importing Android Studio Journeys XML files without requiring conversion to project YAML.
Jul 12, 2026
- v3.0.22
Added support for Nemotron 3.5 and Parakeet TDT v3 models with streaming transcription capabilities. Migrated the desktop app and website to Paraglide i18n to improve internationalization and added translation tests. Enhanced the model selection flow and updated the Wall of Love design.
Jul 12, 2026
- v3.0.21
This release focuses on a major UI overhaul with a redesigned main window and Settings sidebar, alongside new features for dictation control. It also includes performance optimizations for Sona startup and reliability, along with several bug fixes.
Jul 10, 2026
- v3.0.20
This release features a complete rewrite of the Sona component in Rust, resulting in improved performance and maintainability. It introduces a streamlined architecture with faster startup times and simplifies the build process for contributors.
Jul 4, 2026
- v1.90.0
Jul 3, 2026
- v1.89.20
Introduces a pipeline to convert local `SKILL.md` files into verified GEP Recipes, ensuring every procedural step is backed by a real Gene and Capsule with validation evidence.
Jun 29, 2026
- v1.89.19
Enhances the Skill-to-Gene reverse-distillation pipeline with provenance awareness and quality controls to prioritize real execution insights and prevent data leakage.
Jun 29, 2026
- 0.74.0
Improves iOS simulator stability and CLI security by warming up the device to prevent cold-start crashes and masking sensitive API keys in help output.
Jun 25, 2026
- v3.5.1
A maintenance release for `gpt-researcher` that adds new search retrievers and agents while fixing critical bugs related to context management and revision loops.
Jun 23, 2026
- v2.0.4
This release introduces support for YAML configuration files, enabling users to customize settings through a dedicated configuration file rather than relying solely on command-line arguments. It establishes a clear priority system where CLI arguments override YAML files, which in turn override default arguments. Users can easily switch between different config files or run the application without a config file using the new `-c` flag.
Jun 20, 2026
- autogpt-platform-beta-v0.6.63
A comprehensive update to the AutoGPT platform featuring new frontend capabilities like global search and mobile parity, alongside backend improvements for the Copilot bot.
Jun 10, 2026
- v2.0.3
Introduces a new Text-to-Speech (TTS) backend and a comprehensive Web UI for voice management, enabling voice cloning and synthesis testing.
Jun 7, 2026
- v2.0.2
This release focuses on security updates and container runtime compatibility, specifically addressing CVEs and adjusting Vault container capabilities.
Jun 5, 2026
- release-35.3.1v35.3.1
A patch release addressing fixes introduced in version 35.3.0, accessible via the official AG Grid changelog.
Jun 2, 2026