morphik-core Releases

This release introduces comprehensive built-in observability and a pipeline architecture rewrite, enabling granular per-stage latency tracking and declarative provider configuration without external tooling.

Apr 1, 2026

Introduces pluggable Voice Activity Detection (VAD) and End-of-Speech (EOS) engines, comprehensive testing coverage, and infrastructure upgrades.

Mar 17, 2026

A minor bug fix update for the link tool that allows users to open formatted links in a new window when holding the Ctrl key.

Mar 11, 2026

Addresses a stability issue in the inline link tool where selection change events were firing unexpectedly.

Mar 4, 2026

Enhances PDF parsing capabilities with improved table extraction, support for wide tables, and fixes for partially numbered lists.

Feb 20, 2026

Fixes an interaction bug where applying bold formatting to linked text would incorrectly remove the link.

Feb 18, 2026

This release focuses on improving the functionality of the LinkInlineTool by refining the unlink behavior based on the current input state. It also serves as a welcome opportunity to acknowledge a new contributor to the project.

Feb 12, 2026

Feb 7, 2026

Dec 26, 2025

This release enhances CLI flexibility by adding a revision option, introduces a safe transform method for data calibration, and updates dependencies to support newer versions of Lightning and pyannote-database.

Dec 7, 2025

This is a maintenance release focused on security, bumping `mammoth` and `pdfminer.six` to address specific vulnerabilities.

Dec 1, 2025

This release focuses on stability and feature expansion, including pinning ONNX Runtime on Windows, enabling plugin configuration via environment variables, and fixing parsing issues for linked images in DOCX and PPTX shapes.

Aug 26, 2025

Minimal release notes: "introducing transfer (REFER) support to sip-client"

Aug 15, 2025

This release introduces support for Apple Silicon (MPS), updates model weights to safetensors, adds optional CFG and Min_P sampler support, and improves voice conversion examples while fixing CUDA errors.

Jun 13, 2025

This release adds support for advanced AI models including OpenAI o1 and gemini-2.0-flash-exp. It also introduces infrastructure improvements like Redis Sentinel and Cluster support, alongside balance query features for multiple providers.

Feb 2, 2025

This release focuses on bug fixes, improved user experience, and system compatibility updates for Windows, MacOS, and Linux. Key improvements include a browser extension control panel, Docker mode enhancements, and fixes for XPath and encoding errors.

Jan 1, 2025

A substantial release adding support for multiple new AI providers including SiliconFlow, Groq, xAI, and Vertex AI. It includes significant functional updates like OIDC support, multipart/form-data format requests, and custom response formats.

Dec 22, 2024

This release focuses on infrastructure improvements, including Cloudflare Worker AI gateway support, SMTP improvements, and support for Claude 3 tool use. It also introduces new model providers like Novita AI and Vertex AI.

Aug 6, 2024

This release introduces support for the GPT-4o model, Doubao (ByteDance), and Tencent V3 API. It also includes fixes for Gemini streaming, SparkDesk function calling, and various UI/UX improvements.

Jun 25, 2024

Adds support for multiple AI providers including AWS Bedrock (Claude 3), Groq (Llama 3), and Cohere. Also introduces Ollama vision model support and Gemini function/tool support.

May 27, 2024