June

June is a personal AI assistant that runs entirely on your local machine — no subscriptions, no cloud, no payments.

"Hello, what are we working on today?"

What you get

Four pieces, each its own open-source component, assembled into one voice agent:

• The brain: Ollama running qwen2.5:7b (or llama3.2) locally. Fast, private, free.
• The mouth: backtalk — hold a key, speak, and June answers through your speakers a second later. Hearing and voice run on free local models (Whisper + Kokoro).
• The face: ai-visualizer — full-screen visualizer that idles, listens, thinks, and speaks in sync with the conversation.
• The memory: ai-memory-vault — persistent memory built on plain Markdown files in Obsidian.

Requirements

• Windows 10/11 (64-bit)
• Ollama installed
• Python 3.11 (installed via uv automatically)
• A microphone and speakers

Setup

1. Start Ollama

Launch Ollama from the Start Menu. Wait for the llama icon to appear in the system tray.

2. Pull the model

ollama pull qwen2.5:7b

This downloads ~4.4 GB. Do it once. (Or use llama3.2 which is already installed).

3. Install dependencies

$env:Path = "C:\Users\$env:USERNAME\.local\bin;$env:Path"
cd backtalk
uv sync --inexact

4. Launch

Double-click start.bat or run:

.\start.bat

Hold the Home key (or say "go hands free") and speak. June answers through your speakers.

Say "goodbye June" to hang up.

Voice console commands

These exact phrases, spoken alone, control the session:

Phrase	Effect
"goodbye June"	Hang up
"clear the session"	Reset conversation history
"switch to the deep model"	Use qwen2.5:7b / larger model
"back to the fast model"	Return to the default model
"go hands free"	Always-listening mic mode
"push to talk mode"	Hold-key mic mode (default)
"usage report"	Spoken token count for this session

June Studio & Control Center

Manage June's persistent memory and LLM brain endpoints in your browser:

• Studio URL: http://127.0.0.1:8790/dashboard (or click 🧠 STUDIO at the bottom right of any visualizer face)
• Persistent Memory: View and edit memory/MEMORY.md and topic notes with a live markdown editor and preview. Changes sync directly to disk and inject into June's working context.
• Multi-Provider LLMs: Switch between Ollama, LM Studio, OpenAI, OpenRouter, Groq, DeepSeek, or custom OpenAI-compatible endpoints with one click.

Configuration

Settings are managed via the web Studio or by editing backtalk/backtalk.json:

{
  "active_provider": "ollama",
  "providers": {
    "ollama": { "name": "Ollama (Local)", "base_url": "http://localhost:11434/v1", "model": "llama3.2:latest", "deep_model": "qwen2.5:7b" },
    "lmstudio": { "name": "LM Studio (Local)", "base_url": "http://localhost:1234/v1", "model": "local-model", "deep_model": "local-model" },
    "openrouter": { "name": "OpenRouter", "base_url": "https://openrouter.ai/api/v1", "api_key": "sk-or-...", "model": "meta-llama/llama-3.3-70b-instruct" },
    "openai": { "name": "OpenAI", "base_url": "https://api.openai.com/v1", "api_key": "sk-...", "model": "gpt-4o-mini" },
    "groq": { "name": "Groq", "base_url": "https://api.groq.com/openai/v1", "api_key": "gsk_...", "model": "llama-3.3-70b-versatile" },
    "deepseek": { "name": "DeepSeek", "base_url": "https://api.deepseek.com/v1", "api_key": "sk-...", "model": "deepseek-chat" }
  },
  "agent_dir": "d:/Personal/Projects/AI Trials/J.U.N.E",
  "name": "June",
  "ptt_key": "home",
  "voice": "bm_lewis",
  "stt_model": "small.en",
  "stt_device": "cpu"
}

June's personality lives in AGENTS.md and persistent memory lives in memory/MEMORY.md.

Troubleshooting

See TROUBLESHOOTING.md for common issues. If something breaks, open a chat and describe the problem — June is built to diagnose and fix her own stack.

License

Code components are licensed under the GNU Affero General Public License v3 or later (AGPL-3.0-or-later). See LICENSE.

Built by Akhil, powered by open-source tools.