FAQ

Frequently Asked Questions (FAQ)

General

What is Mingly?

Mingly is a desktop application that lets you use multiple AI models (Claude, GPT, Gemini, Llama and more) in a single interface. You keep full control over your data and API keys.

Is Mingly free?

The Free tier is permanently free with local models (Ollama). For cloud APIs, team features, and advanced capabilities there are paid plans:

• Free — CHF 0, local models, 3 conversations/day
• Pro — CHF 24/month, cloud APIs, unlimited conversations, AI agents, prompt templates
• Team — CHF 69/user/month, team workspaces, RBAC, audit logs, SSO
• Enterprise — On request, on-premise, compliance, dedicated support

The app stays open source under MIT license. See Pricing for full details.

Which AI models are supported?

• Anthropic: Claude 4.5 Sonnet, Claude 4 Opus, Haiku
• OpenAI: GPT-4o, GPT-4 Turbo, GPT-3.5 Turbo
• Google: Gemini 2.5 Pro, Gemini 2.0 Flash
• Ollama: Any locally available model (Llama, Mistral, Phi, etc.)

Costs

How much does it cost to use?

Costs depend on the AI provider and model. Approximate pricing per 1 million tokens:

Model	Input	Output
Claude Haiku	~$0.25	~$1.25
Claude Sonnet	~$3.00	~$15.00
GPT-4o	~$2.50	~$10.00
GPT-3.5 Turbo	~$0.50	~$1.50
Ollama (local)	Free	Free

Tip: Use the budget feature in Settings to cap your monthly spending.

Can I limit costs?

Yes. Under Settings > Budget, you can set monthly limits. Mingly warns you before the limit is reached.

Privacy

Where is my data stored?

All conversations, settings, and the knowledge base are stored exclusively on your computer. Mingly does not send data to its own servers.

Are my chats sent to AI providers?

Yes, your messages are sent to the respective AI provider (e.g., Anthropic, OpenAI) to generate a response. This is required for the service to work. Exception: With Ollama, everything stays local.

Is Mingly GDPR-compliant?

Mingly includes features for GDPR compliance:

• Data export: Export all your data as a file
• Data deletion: Completely remove all your data
• No telemetry: No usage data is collected
• Sensitive data detection: Warns before accidentally sending passwords

How secure are my API keys?

API keys are stored in the OS secure keychain:

• macOS: Apple Keychain
• Windows: Windows Credential Vault

They are never stored in plain text on disk.

Local AI (Ollama)

How do I use AI without the cloud?

1. Install Ollama
2. Pull a model: ollama pull llama3.2
3. Start Ollama
4. Select "Ollama" as provider in Mingly

All processing happens locally on your computer — no data leaves your device.

What hardware do I need for local models?

• Minimum: 8 GB RAM (for small models like Phi)
• Recommended: 16 GB RAM, Apple M1/M2/M3 or dedicated GPU
• Large models (70B): 32+ GB RAM

Server Mode

When do I need server mode?

Server mode is useful for:

• Teams: Multiple people share a single Mingly instance
• Centralization: Manage API keys in one place
• Integration: Connect other applications via the REST API
• Headless: Run on a server without a desktop UI

How do I set up the server?

The easiest method is Docker:

docker compose up -d

Or manually:

npm run build:server
npm run start:server

The server is then available at http://localhost:3939.

Can I secure the server?

Yes, with API key authentication:

1. Set MINGLY_REQUIRE_AUTH=true
2. Define MINGLY_API_KEY=YourSecretKey
3. All requests must include the header Authorization: Bearer YourSecretKey

Troubleshooting

The app won't start

1. Make sure system requirements are met
2. Try reinstalling the app
3. For source installation: rm -rf node_modules && npm install

Responses are slow

• Check your internet connection
• Choose a faster model (e.g., Haiku instead of Opus)
• For Ollama: Make sure enough RAM is available

The knowledge base finds nothing

• Check if the folder path is correct
• Make sure files are in a supported format
• Wait for indexing to complete

---

Back to: Installation | Configuration | Usage | Pricing