PaddleSpeech Releases

Enhanced CSV parsing error diagnostics to help users debug silent failures and improved Windows test compatibility.

Jan 17, 2026

Introduced CSV export functionality for spreadsheets and dynamic transaction description transformation capabilities.

Jan 14, 2026

Added support for directory and glob patterns in CSV sources, improved European CSV handling, and fixed tag collection logic.

Jan 10, 2026

Introduces the new primary command `tally up`, adds a Report Diff feature to track changes, and adds experimental supplemental data sources for enriching transactions.

Jan 6, 2026

Maintenance release updating dependencies to address security vulnerabilities in mammoth and pdfminer.six.

Dec 1, 2025

Adds support for StrictJSON Schema with OpenAI and Gemini models, focusing on generating Pydantic models for structured outputs, and refactors internal YAML parsing utilities.

Nov 20, 2025

Enhances parse_yaml robustness for handling special characters and complex structures, adds image parsing capabilities for compatible providers, and updates Pydantic to v2.12.4.

Nov 7, 2025

Adds remote webhook notifications, pause and resume capabilities for packet captures, and includes Czech translation and Windows ARM64 support.

Nov 4, 2025

A maintenance release that fixes a regression regarding comment nesting in code fences and nunjucks, and updates the package.json repository field format.

Oct 31, 2025

Focuses on performance optimizations including cache fixes, reduced overhead, and fixes for Table of Contents generation, alongside dependency updates.

Oct 26, 2025

This release focuses on dropping support for older Node.js versions (16 and 18) while introducing significant performance optimizations like caching and faster text-tables. It also includes new features for permalink customization and URL validation, alongside major dependency upgrades.

Sep 15, 2025

A massive update involving 79k lines of code added primarily by AI agents. It introduces a new strong static typed design, database implementations, and enhanced autonomous agent capabilities.

Jun 9, 2025

Includes security updates, Docker version bumps, and a breaking change where the Console feature is disabled by default, alongside translation updates and UI fixes.

Mar 30, 2025

Major version adaptation release upgrading PaddleSpeech from Paddle 2.5.1 to Paddle 3.0.0-beta, with 80+ models adapted and 20+ dynamic-to-static models re-exported using PIR + predictor approach. Also introduces AudioTools toolkit for DAC (Descript-Audio-Codec) training with new loss functions.

Mar 5, 2025

This release introduces several new AI models and enhancements to existing LLM integrations, specifically focusing on advanced reasoning models and new provider support.

Feb 26, 2025

A maintenance and refactoring release that focuses on UI improvements, agent automation for code reviews, and a rebranding to 'TypedAI'.

Feb 17, 2025

A significant feature release adding advanced reasoning models (OpenAI o1/o3-mini), multi-agent implementations (CePO, debate), and UI enhancements like keyboard shortcuts and caching.

Feb 3, 2025

Expands model support with Gemini 2.0 Flash Thinking and DeepSeek v3, adds chat attachments and PDF support, and improves SWE agent project handling.

Jan 6, 2025

A maintenance release for the Hexo static site generator focusing on performance improvements, bug fixes, and tooling upgrades.

Jul 1, 2024

New ASR models and TTS improvements including WavLM, HuBERT, and Squeezeformer for speech recognition, plus SVS (Singing Voice Synthesis) examples with Opencpop dataset.

Jun 27, 2024