Skip to content
toolbay

Toolbay News ·

Handy nears 1.0, OpenWhispr adds team words

Free dictation app Handy adds audio-file transcription, OpenWhispr shares team dictionaries, and Open WebUI and Ollama ship new releases for local AI.

Edited by Orion Jones · How we report

  1. Handy 0.10.0 can transcribe audio files

    The free, open-source dictation app Handy released version 0.10.0 on November 2. It adds transcription of audio files, automatic language detection for all the transcription models it supports, and a new onboarding flow. The maintainer says the app is roughly feature complete and that coming releases will focus on stability on the way to V1.

    Our view A dictation app that runs on your own machine is now close to a stable 1.0, which matters if you want offline dictation without a subscription.

    Source: Handy on GitHubOpen-source dictation apps

  2. OpenWhispr 1.11.0 adds a shared team dictionary

    OpenWhispr 1.11.0 lets Business and Enterprise members share a team dictionary that applies to dictation and meeting notes, with your own words still taking priority. It also adds saved Skills for recurring work and a workspace switcher. Whisper now checks that CUDA or Vulkan actually started and falls back to the CPU if not.

    Our view The GPU fallback is the change most useful to individual Windows users; the dictionary is aimed at teams.

    Source: OpenWhispr on GitHubVoice dictation for Windows

  3. Open WebUI 0.13.3 lets agents run on a schedule

    Open WebUI 0.13.3 lets the automation editor pick a terminal and working folder, then choose a coding agent available there. Scheduled runs use that saved terminal and folder and check they are still available before starting. You can also change the terminal or folder partway through a chat.

    Our view Scheduled agent runs make this self-hosted chat app more of a work tool, so check which terminals it can reach before turning them on.

    Source: Open WebUI on GitHub

  4. Ollama 0.40.15 adds a 9B code model and a launcher

    Ollama 0.40.15 adds support for Xiaomi's mimo-v2.6, a 9B model built for code and agent workflows, and for AutoTrust decision models. Recurrent and vision GGUF models can now serve more than one request at a time, and running ollama with no arguments opens a welcome launcher.

    Our view Small local models keep getting easier to try; the launcher helps newcomers who do not want to learn commands first.

    Source: Ollama on GitHub