Paddle Releases

This release focuses on fixing critical bugs related to document processing stability and Unicode handling. It addresses data loss issues in DOCX tables and improves PDF text extraction by implementing surrogate pair restoration. Additionally, the project welcomes a new contributor.

Aug 14, 2026

This release introduces a new FunASR backend and fixes critical stability issues with qwen3 diarization. It also enhances Docker image security with SBOM attestations and improves error handling for invalid session parameters.

Aug 1, 2026

This release introduces the AlignAtt translation backend for streaming LLM translation, refreshes benchmark figures to correct previous inaccuracies, and improves error handling to fail loudly when the ASR backend produces no output.

Jul 11, 2026

This release adds Qwen3-ASR backends, introduces prebuilt Docker images, adds API token authentication, and fixes various bugs related to session stability, timeouts, and memory usage.

Jul 9, 2026

This release updates the FunASR llama.cpp runtime binaries, focusing on improving CPU compatibility and performance. It introduces a specific AVX2 build for faster execution on compatible hardware while maintaining a conservative default build to prevent crashes.

Jun 29, 2026

This release is a portable rebuild of the runtime binaries designed to fix stability issues. It disables aggressive CPU optimizations found in the runner environment to ensure the binaries work on a wider range of generic CPUs without crashing.

Jun 28, 2026

This release introduces the ability to run SenseVoice (and Paraformer / Fun-ASR-Nano) locally using the llama.cpp/GGUF runtime. It emphasizes on-device ASR capabilities and introduces a new quantization option for smaller model sizes.

Jun 21, 2026

A runtime release providing prebuilt, self-contained binaries to run SenseVoice and Paraformer locally via llama.cpp, featuring on-device ASR capabilities and built-in VAD without requiring Python or build tools.

Jun 21, 2026

This is a maintenance release focused on dependency updates and resolving a critical bug that prevented users from uploading docx files in the guest signing flow.

Jun 17, 2026

A feature-rich release introducing attachments widgets and API support, alongside significant improvements to user authorization, webhook handling, and validation for draft documents and templates.

Jun 17, 2026

This release focuses on core audio processing improvements, specifically normalizing Chinese language codes for transcription and fixing timestamp drift issues in Voice Activity Detection (VAD). It also optimizes CUDA memory management for performance.

Jun 4, 2026

This release introduces role-based signing workflows, offline signing capabilities, and enhanced API features for PDF handling and document tracking. It also improves the user interface and fixes critical bugs related to templates and mobile views.

Jun 2, 2026

The first official release of the SenseVoiceSmall checkpoint introduces multilingual speech understanding capabilities, supporting ASR and language identification for five languages while also detecting emotions and audio events.

May 25, 2026

This release introduces a new Parquet reader input type, NLP streaming capabilities, and client-side JSON validation. It also addresses security vulnerabilities by bumping the nltk dependency and fixes documentation links.

Apr 13, 2026

A patch release focused exclusively on security, resolving a vulnerability in the NLTK library.

Mar 27, 2026

This release prioritizes security by pinning a compromised dependency and optimizes vector storage configuration based on embedding models.

Mar 24, 2026

This release enhances the e-signature platform with new UI options, improved widget functionality, and API updates including BCC support.

Mar 16, 2026

This release updates the core ASR backend to Qwen3 Beta and introduces optimized WebSocket protocols for bandwidth efficiency in session tracking.

Mar 12, 2026

This release announces the MiroFish swarm intelligence engine, a tool for simulating future scenarios based on input data, with comprehensive deployment guides.

Mar 7, 2026

This release focuses on data cleaning and performance improvements by removing phantom entities from relationships and introducing NLP streaming capabilities.

Mar 6, 2026