Keep your Z.ai / BigModel API keys alive and monitored — silently, from your system tray.

GLM Tray is a lightweight, native desktop app that sits in your system tray and watches your API keys so you never hit an unexpected quota wall or stale key. Manage up to 4 keys, visualise live quota usage, and automate keep-alive pings — all without leaving your workflow.

• 🔑 Multi-key dashboard — Monitor up to 4 Z.ai / BigModel API keys side by side
• 📊 Live quota tracking — Token limits, request counts, and model-level breakdowns at a glance
• 💓 Keep-alive scheduler — Three flexible modes keep keys warm automatically: Interval, Specific Times, or After Reset
• ✅ Smart wake confirmation — Validates success via quota delta, retries silently on failure
• 🌐 Dual platform support — Works with both api.z.ai and open.bigmodel.cn endpoints
• 🔔 Auto-update notifications — Stay current with in-app update prompts
• 🗓 JSONL audit logging — Optional, filterable logs with flow_id and phase fields
• ⚙️ Global app settings — Configure shared quota URL, LLM URL, log path, and retention from one place

Grab the latest release for your platform from the Releases page →

Windows

1. Download and run the .exe installer
2. Follow the installation wizard

macOS

1. Download the .dmg file
2. Drag GLM Tray to Applications
3. On first launch: right-click → Open (or allow in System Preferences → Privacy & Security)

Linux

chmod +x glm-tray_*.AppImage
./glm-tray_*.AppImage

1. Launch the app — it appears silently in your system tray
2. Click the tray icon — opens the main window
3. Add your API key — paste your Z.ai or BigModel key into Slot 1
4. Enable polling — toggle on Enable polling to begin monitoring
5. Check your usage — live stats appear instantly on the dashboard

Stay ahead of your limits with real-time usage data:

• Token quota consumption and limits
• Request counts and model-level breakdowns (24-hour window)
• Tool usage statistics
• Visual indicators directly in the tray icon

Three scheduling modes prevent stale keys — mix and match as needed:

Wake requests are verified — not just sent:

• After a wake, the slot is marked wake_pending and an immediate quota poll fires
• If quota shows a valid nextResetTime advance → ✅ confirmed, wake_pending clears
• If quota doesn't confirm → retry every minute for the configured wake_quota_retry_window_minutes
• After the window, a forced retry is attempted
• Persistent failures increment wake_consecutive_errors; once the threshold is reached, the slot is temporarily auto-disabled for wake

Enable structured logging to debug API interactions:

• Daily .jsonl log files with full request/response data
• flow_id ties each request/response pair together
• phase field: request, response, error, event
• Scheduler events logged: wake pending, retry windows, task start/stop
• Default path: {app_data}/logs/ — override with log_directory in Global Settings

Access via the gear icon on the home page:

Settings are stored in the platform-standard application data folder:

For BigModel, replace api.z.ai with open.bigmodel.cn.

• Rust 1.70+
• Node.js 18+

sudo apt-get install -y \
  libwebkit2gtk-4.1-dev build-essential curl wget file \
  libssl-dev libgtk-3-dev libayatana-appindicator3-dev \
  librsvg2-dev patchelf pkg-config libsoup-3.0-dev \
  javascriptcoregtk-4.1 libjavascriptcoregtk-4.1-dev

xcode-select --install

• Visual Studio Build Tools (Desktop development with C++)
• WebView2 Runtime (usually pre-installed on Windows 11)

git clone https://github.com/kiwina/glm-tray.git
cd glm-tray
npm install
npm run tauri dev

npm run tauri build

Built installers land in src-tauri/target/release/bundle/.

Test wake functionality without hitting production APIs:

node docs/mock-server.cjs   # Start mock server

Then enable Debug Mode in Global Settings → Developer section. All API calls route to the mock server.

See docs/DEBUGGING.md for full documentation.

src/
  main.ts              # Frontend entry (Vue 3 + Pinia)
  styles.css           # DaisyUI + Tailwind CSS 4

src-tauri/src/
  lib.rs               # Tauri setup, commands, state
  config.rs            # Config load/save with migration
  api_client.rs        # HTTP client for API calls
  scheduler.rs         # Background polling scheduler
  tray.rs              # System tray management
  models.rs            # Shared data structures
  update_checker.rs    # Auto-update checker
  file_logger.rs       # JSONL logging module

• v0.0.4 — Vue 3 rewrite, multi-key dashboard, global settings, JSONL logging, auto-updater
• v0.0.3 — Keep-alive scheduler with wake confirmation and retry logic
• v0.0.2 — Dual platform support (Z.ai + BigModel)
• v0.0.1 — Initial release

MIT

This software is not affiliated with, endorsed by, or sponsored by Z.ai, BigModel, or any of their subsidiaries.

"Z.ai" and "BigModel" are trademarks of their respective owners. This is an independent, community-developed tool for personal API key management. Use at your own risk.

The software is provided "as is", without warranty of any kind, express or implied. The authors are not liable for any damages arising from the use of this software.