3.4 AI Service Provider Pricing Information
Billing Method
Token Billing:
- Most services charge by Token
- 1 Token ≈ 0.75 English words (varies by language)
- Cost = Input Tokens + Output Tokens
Example:
- User message: "Hello, how's the weather today?" (about 15 Tokens)
- AI reply: "The weather is sunny today, temperature is comfortable..." (about 50 Tokens)
- Total: 65 Tokens
FAQ
Q: Can I use it for free?
A:
- MiniTavern provides free quota for new users, allowing you to start conversations without configuration
- You can also choose providers with free quotas (such as Nvidia, Google Gemini)
- Or use Other LLM to configure Ollama local deployment (completely free, but requires hardware)
Q: Which service is cheapest?
A:
- Cheapest: DeepSeek
- Highest free quota: Google Gemini
- Local free: Ollama
Q: Will API Keys expire?
A:
- API Keys themselves usually don't expire
- But you need to recharge when account balance runs out
- May be reclaimed by service provider if unused for long periods
Q: Can I use multiple services simultaneously?
A:
- Yes! MiniTavern supports configuring multiple APIs
- Different characters can use different models
Q: Will costs be high?
A:
- Daily use is usually not expensive
- You can set monthly budgets to control costs
Q: Is data secure?
A:
- MiniTavern does not store your conversations
- API Keys are only saved locally
- Conversations are sent to AI service providers for processing
- Using Ollama can achieve complete localization without data security concerns