filestash Releases

Version 0.4.0 introduces a range of enhancements to the drawing application, including touch device support, new export formats (SVG, PNG), and UI improvements like a 'More Options' dropdown and Vietnamese translations.

Apr 19, 2026

This release introduces explicit tag dependency semantics using AnyOf and AllOf modifiers, alongside resolution tracking and event loop health monitoring. It also refines the core relationship entity kinds and fixes several issues related to dependency propagation and SDK startup.

Apr 14, 2026

This release prioritizes security hardening with fixes for Mass Assignment and IDOR vulnerabilities across multiple endpoints. It also adds new integrations for Pipedream and Browserless MCP, along with significant updates to the Agentflow UI and state management.

Apr 14, 2026

Flowise 3.1.1 introduces a Rich Text editor for content editing, adds support for new OpenAI models (gpt-5.4-mini and gpt-5.4-nano), and enhances AgentFlow with new input types and load methods. The release also includes various bug fixes, dependency updates, and improvements to the UI experience.

Mar 23, 2026

This release focuses on security hardening by enabling HTTP security checks by default, adding Azure Blob Storage support, and introducing the new `@flowiseai/agentflow` SDK. It also includes core reasoning support, AWS STS integration, and various UI/UX improvements.

Mar 16, 2026

This release introduces per-agent planners, an advanced tagging system for guidelines, and new SDK methods for session and customer management. It also upgrades the MCP service and fixes various stability and security issues.

Mar 15, 2026

This release introduces new parameters for creating guidelines, observations, and journeys, alongside an alias for filtering. It addresses several stability issues, including a server hang during setup and improved message handling. Additionally, it deprecates the `attach_tool()` method in favor of a new parameterized approach.

Feb 17, 2026

This release introduces advanced lifecycle management for journeys and guidelines, featuring automatic label propagation, new hooks, and streaming output, while fixing critical agent logic and websocket issues.

Feb 9, 2026

This release introduces a Flask-based Web UI, EPUB 3 packaging pipeline, and GPU-accelerated TTS support, alongside a major codebase reorganization to support both desktop and web interfaces.

Feb 6, 2026

This release introduces support for the `uv` package manager to speed up Abogen installation and adds a new option in the queue manager to override settings. It also resolves several critical bugs related to Linux compatibility, dependency management, and text processing accuracy.

Dec 10, 2025

This release enhances subtitle generation capabilities by introducing support for non-English languages and improved sentence segmentation via spaCy. It also adds offline functionality through model pre-downloading and includes performance optimizations. Several bugs related to text truncation and regex handling have been resolved.

Nov 28, 2025

This release introduces the ability to generate audio from subtitle files (.srt, .ass, .vtt) and adds support for embedding cover images in M4B files. It also addresses critical Windows DLL initialization errors and CUDA availability issues. Furthermore, improvements to code and documentation have been included.

Nov 18, 2025

This release focuses on UI improvements, including upgrading the Abogen interface to PyQt6 and adding tooltips for book handler options. It also introduces new file opening actions in the queue manager and visual feedback like loading animations.

Oct 27, 2025

This release introduces a new eraser tool, expands internationalization support to include Russian and Arabic, and fixes various hotkey conflicts. It also upgrades the core Plait library and addresses several security and UI stability issues.

Sep 14, 2025

Introduces a new graphical user interface (GUI) that enables precise control over voice synthesis, allowing users to adjust pitch and duration values and swap voices across 7000+ languages.

Oct 7, 2024

A minor release focusing on significant audio quality improvements by retraining the vocoder end-to-end on synthetic spectrograms rather than ground-truth ones, resulting in better quality without sacrificing inference speed.

Sep 22, 2024

Enhances the universal TTS model for 7000 languages with stochastic prosody prediction, expanded IPA modifier support, additional pretraining languages, and improved language similarity visualization.

Jul 25, 2024

A major release introducing a massively multilingual Text-to-Speech system capable of synthesizing speech in over 7000 languages, featuring watermarking, improved quality, and extrapolation capabilities derived from a 462-language model.

Jun 10, 2024

Adds the ability to control the emotional tone of synthesized speech by conditioning the model on emotional prompts during training and inference.

Jun 10, 2024