archify Releases

Updates the CodeChunker resulting in significant performance improvements, reducing index time by 44.5% and improving latency across the board.

May 13, 2026

Minimal release notes: "* Update dependencies in https://github.com/fregante/GhostText/pull/349 **Full Changelog**: https://github.com/freg"

May 12, 2026

A patch release focused on documentation updates and fixing UI/UX issues, including canvas rendering and export menu clarity.

Apr 15, 2026

This release improves raster export quality by implementing native 4× scaling and removes the manual scale selector.

Apr 15, 2026

This patch release fixes a visual bug where security group labels were overlapping with inner components. It adjusts the template dimensions and adds documentation to ensure consistent spacing for future diagrams.

Apr 15, 2026

ECC v1.10.0 focuses on improving harness reliability, cross-platform compatibility, and operator workflow ergonomics. It also previews the upcoming ECC 2.0 operating system.

Apr 5, 2026

A maintenance release focused on ensuring compatibility with Torch 2.8 by pinning specific dependency versions to prevent conflicts.

Apr 1, 2026

Major upstream migration to more stable and reliable tools, resolving deprecated dependencies and high deployment barriers.

Mar 31, 2026

A backport of timestamp fixes from v3.8.4 that restores the ability to generate timestamps for words containing digits, symbols, or foreign scripts.

Mar 25, 2026

Identical to v3.7.9, this release is a backport of timestamp fixes from v3.8.4 to restore wildcard emission column support for special characters.

Mar 25, 2026

This release adds significant new capabilities including GPU control, a new HuggingFace NER recognizer, recognizers for Nigeria and the UK (postcode, passport, vehicle), and ONNX Runtime support.

Mar 18, 2026

Adds a new WeChat Official Account channel and fixes critical Windows compatibility issues.

Mar 4, 2026

Focuses on global recognizers (Thailand, India, Korea) and infrastructure improvements like Docker healthchecks, GPU optimizations, and the integration of language models (LangExtract).

Feb 12, 2026

This release addresses critical stability issues by fixing a healthcheck race condition that caused Gluetun to hang, resolving Wireguard kernel detection problems, and updating ProtonVPN settings to support UDP port 51820.

Feb 11, 2026

Early development milestones including Dockerization, MPS support for Apple Silicon, and integration of multiple STT/TTS models such as Paraformer and ChatTTS.

Feb 6, 2026

The latest update enhances LLM integration by enabling automatic model loading and connection testing. It also adds user guidance for self-hosted workflows and fixes a parameter passing bug in digital human broadcasting.

Jan 27, 2026

The Pixelle-Video integration package introduces an action migration module and local workflow support for digital humans and image-to-video generation. It resolves TTS instability by locking the edge-tts version to ensure reliable audio generation.

Jan 26, 2026

This release expands the platform's capabilities with a digital human video generation pipeline and an Audio-Visual pipeline for image-to-video. It adds multi-language TTS support and optimizes deployment for machines with 48GB of VRAM.

Jan 14, 2026

The update adds the Nano Banana image generation workflow and enhances ComfyUI API integration with customizable parameters. It also optimizes LLM processing and introduces configuration hot-reload capabilities.

Dec 28, 2025

This release focuses on DNS improvements, including support for DoH and local network resolution, alongside a reworked healthcheck system and updates to built-in VPN server data.

Dec 25, 2025