OmniRoute Releases

This release backports the word-level timestamp fixes from v3.8.4 to restore functionality for words containing special characters or digits.

Mar 25, 2026

Added support for 10 AI coding agents (Cline, Copilot, OpenClaw, Amp, Antigravity, Gemini) within the `mx init` command, allowing users to configure specific rules and instructions for each agent.

Mar 24, 2026

Introduced a per-project merged search index that consolidates package data into a single FTS5 query, significantly reducing search time from 310ms to 70ms and improving efficiency during package management.

Mar 23, 2026

Updated integration instructions for all AI assistants to include missing commands, positional argument warnings, and result count guidance, while removing outdated flags and clarifying that semantic reranking is now default.

Mar 23, 2026

Overhauled reranking performance by 3.3x by implementing a single pass reranking strategy, using pre-optimized ORT models, and enabling zero-copy memory-mapped file access with multi-threaded inference.

Mar 23, 2026

Replaced the HuggingFace tokenizers library with `tokie` to achieve massive performance improvements in reranking, reducing initialization time and dependencies while maintaining 100% accuracy.

Mar 23, 2026

A major update focusing on Voice AI training and LLM capabilities, specifically adding VibeVoice LoRA finetuning and Qwen3.5 support. It also includes significant UI refinements, Python 3.11 as the default, and audio denoising options.

Mar 5, 2026

This release prioritizes performance optimization by integrating Faster-Qwen3-TTS for CUDA graph acceleration and introduces multi-GPU support to allow subsystems to run on separate hardware.

Mar 1, 2026

This update enhances the utility of the tools by adding a centralized Prompt Hub for quick access to saved prompts and introducing Ollama as a new LLM backend. It also adds a batch generation feature to split text by paragraph.

Feb 22, 2026

This release focuses on user experience improvements, introducing multiple output formats, embedded metadata for self-contained audio files, and robust persistent settings that survive application restarts.

Feb 14, 2026

Focuses on stability improvements by fixing a critical healthcheck race condition that caused the container to hang, along with updates to Wireguard detection and ProtonVPN configurations.

Feb 11, 2026

An early foundation release focusing on Dockerization, MPS support, and the integration of various speech backends like Paraformer and ChatTTS.

Feb 6, 2026

Focuses on improving automation reliability by refining form filling logic and key press handling to prevent misidentification and improve data entry.

Jan 12, 2026

A major feature release introducing a completely reworked healthcheck system with multiple check types, DNS over HTTPS support, and enhanced port forwarding capabilities, alongside significant dependency updates.

Dec 25, 2025

A maintenance release fixing regressions in Wireguard endpoint IP handling and resolving updater issues for various VPN providers, ensuring data integrity and connection stability.

Dec 24, 2025

This release focuses on performance optimization, configuration improvements, and bug fixes. It introduces connection pooling for DNS over TLS, new configuration options for health checks and rebind protection, and fixes various logging and error handling issues.

Dec 22, 2025

Minimal release notes: "- update the inference pipeline and models to support unlimite audio length - update the models cards"

Dec 22, 2025

This release addresses several stability and compatibility issues in scrcpy, including permission handling after Android upgrades, state restoration failures, and startup issues on specific devices.

Dec 17, 2025

Dec 12, 2025

Updates the underlying framework and introduces a new dynamic agent for better context management and tracing capabilities.

Nov 15, 2025