Releases·Tool Release
Vellium v1.1.0 adds live voice mode with local STT/TTS and easier llama.cpp setup
Vellium, an open-source local-first desktop app for AI chat, roleplay, and long-form writing, released v1.1.0. The update centers on live voice: microphone input, local or Whisper-compatible speech recognition, and streaming TTS within one voice interface. It also simplifies local model setup via llama.cpp, relevant to practitioners running fully offline assistants.
45 engagement1 sourceTue, Sep 1, 2026, 10:45 AM
AI Overview
- Live mode supports microphone input, attachments, screen context, and standard chat tools inside the same voice interface
- Local speech recognition uses Whisper Large v3 Turbo Q5_0, installable and configurable directly in the app
- TeraTTSv2 provides English and Russian voices with realtime streaming playback
- Setup for llama.cpp local models was simplified in recent updates
Related
Social·Community
Community shares local LLM chat app preferences
A Reddit user asked what chat applications the community uses for local LLMs, sparking discussion about tools like LMStudio, llama-server, Open WebUI, and SillyTavern. The thread highlights the trade-offs between developer access and feature richness.
A Reddit user asked what chat applications the community uses for local LLMs, sparking discussion about tools like LMStudio, llama-server, Open WebUI, and SillyTavern. The thread highlights the trade-offs between developer access and feature richness.
53 engagement1 sourceredditWed, Jul 22, 2026, 03:50 PM
Releases·Tool Release
Vellium v1.0.0 released: security hardening, wallpaper-based themes, JSON chat export and a major desktop stability pass
Vellium, a local-first desktop workspace for writing and roleplay with local LLMs, has reached its first stable release v1.0.0. The release focuses on security hardening, desktop stability, and new features like wallpaper-based themes and JSON chat export.
Vellium, a local-first desktop workspace for writing and roleplay with local LLMs, has reached its first stable release v1.0.0. The release focuses on security hardening, desktop stability, and new features like wallpaper-based themes and JSON chat export.
1 engagement1 sourceredditSun, Jul 12, 2026, 08:54 AM
Products·launched
Local LLM coding harness with just-in-time code review
An open-source harness for running local LLMs, built with Qwen 3.x 27B, that replaces OpenCode. It features just-in-time code review before tool calls and supports three-way chat between agents and users in sub-agent threads. It works on Windows and Linux, is free, has no telemetry, and is designed for coding and mixed workflows.
An open-source harness for running local LLMs, built with Qwen 3.x 27B, that replaces OpenCode. It features just-in-time code review before tool calls and supports three-way chat between agents and users in sub-agent threads. It works on Windows and Linux, is free, has no telemetry, and is designed for coding and mixed workflows.
209 engagement2 sourcesredditThu, Aug 27, 2026, 02:40 PM
Products·launched
ChungusHub: Local-first LLM frontend for roleplay with tree chat
ChungusHub is a local-first, privacy-focused LLM frontend for roleplay. It uses SillyTavern's formats for characters, lorebooks, and personas, allowing easy import of existing libraries. Its chat is a tree structure, enabling branching edits and regenerations, with a story map for navigation.
ChungusHub is a local-first, privacy-focused LLM frontend for roleplay. It uses SillyTavern's formats for characters, lorebooks, and personas, allowing easy import of existing libraries. Its chat is a tree structure, enabling branching edits and regenerations, with a story map for navigation.
29 engagement1 sourcegithubWed, Aug 19, 2026, 12:01 PM
Products·launched
Ryza AI Revive — offline fan-made Ryza AI companion with bring-your-own LLM and TTS
A static web app with Windows and Android shells that acts as an offline AI companion themed on Ryza. Users supply their own LLM and TTS API keys in settings, so nothing connects to official servers. It solves the problem of a private, local-configurable character companion for fans of the game series.
A static web app with Windows and Android shells that acts as an offline AI companion themed on Ryza. Users supply their own LLM and TTS API keys in settings, so nothing connects to official servers. It solves the problem of a private, local-configurable character companion for fans of the game series.
46 engagement1 sourcegithubWed, Sep 2, 2026, 12:52 AM