Practical overview
When it fits
Use June when you want a configurable command-line chatbot that combines a locally run Ollama language model with speech recognition and optional synthesized speech.
Who it is for
- People who want to interact with a chatbot by speaking and hearing synthesized replies on their own machine.
- Developers who want to configure a local assistant’s language, speech-recognition and speech-synthesis models.
- Users who want text-only interaction with the option to disable speech components.
Documented uses
- Hold a conversation by speaking into a microphone and receiving text and synthesized audio responses.
- Submit spoken input and receive text responses, or provide text and receive audio output.
- Select model identifiers and Torch devices through a JSON configuration file.
- Disable speech recognition or synthesis to use text-based interaction.
Requirements & limits
The project is an open-source GitHub repository identified as MIT-licensed. Its documented setup requires local installation, Ollama and Python 3.10 or later.
- June separates language generation, speech recognition and synthesis across Ollama, Hugging Face Transformers and Coqui TTS.
- The default configuration uses Llama 3.1 8B Instruct, Whisper Small English and Glow-TTS.
- Configuration supports model IDs and Torch device choices, including CPU, CUDA and MPS examples.
- The README says enabled chat history consumes more processing power.
- The documented voice-input flow requires no wake command and asks the user to remain silent for three seconds after speaking.
Check current requirements in the official documentation. Source-based guidance does not mean OpenViST has run or independently tested this project.
Install June and launch its default assistant
Install June from its GitHub repository, fetch its default Ollama model and start the command-line assistant.
Before you start
- Ollama
- Python 3.10 or later with pip
- GNU/Linux: Python development package and PortAudio development package; macOS: PortAudio; Windows: Microsoft Visual C++ Build Tools 14.0 or later
Steps
Install June using pip from the GitHub repository, or clone the repository and run pip install . from its directory.
Pull the default language model with ollama pull llama3.1:8b-instruct-q4_0.
Start the default assistant with june-va.
Expected outcome
June starts with the documented default model combination: Llama 3.1 8B Instruct, Whisper Small English and Glow-TTS.
Guide limitations
- The listed platform-specific development prerequisites vary by operating system.
- The default voice interaction uses microphone input and synthesized audio output.
Configure June for text-only interaction
Use June without its speech-recognition and audio-synthesis components by setting their configuration entries to null.
Before you start
- June installed with its documented prerequisites
- A JSON configuration file
Steps
Create a JSON configuration file and set the stt entry to null to disable speech recognition.
Set the tts entry to null as well to disable audio synthesis.
Pass the configuration file to June with june-va --config path/to/config.json.
Expected outcome
The README describes disabling both speech components to use the assistant in text mode.
Guide limitations
- The guide covers configuration, not verification of a completed session.
Sources & review
Source record details
- Source availability
- Active
- Sources checked
- 15 October 2026
- Original website title
- mezbaul-h/june · Official homepage title · retained in its source language