[Feature]: Quick Start with One API Key + Embedded DB #161
Description
Describe your use case
Lower First-Time Usage Barrier: One-Click API Key Integration with Popular LLMs/Embedders + Embedded Database Out of the Box
Description
To improve the onboarding experience for new users, we propose streamlining the initial setup process into a “configure-and-go” minimal mode. The following enhancements will significantly reduce the entry barrier:
Describe the solution you'd like
1. One-Click Configuration for Major LLM API Keys
Users simply provide a single API key (e.g., from OpenAI, Qwen, etc.), and the system automatically detects and enables the corresponding:
- LLM (for reasoning, generation, etc.)
- Embedder (for vectorization)
2. Preconfigured, High-Quality Default Model Pairs
For each supported platform (e.g., OpenAI, Qwen), the system includes validated default model combinations, such as:
- OpenAI:
gpt-4o(LLM) +text-embedding-3-large(Embedder) - Qwen:
qwen-max(LLM) +text-embedding-v3(Embedder)
Users don’t need to manually select models or understand technical differences—high-performance defaults work out of the box.
3. Embedded Database by Default (e.g., SQLite)
If no explicit configuration is provided, the system automatically uses a lightweight embedded database (e.g., SQLite or SeekDB in local mode) to store memories and vector data—eliminating the need to deploy external dependencies.
Expected Outcomes
- New users: Complete full deployment and run all features within 5 minutes using just one API key.
- Developers / demo scenarios: Launch demos instantly without managing infrastructure.
- Reduced cognitive load: Complex configuration options are hidden by default; the preset path is robust and sufficient for most use cases, while advanced settings remain overrideable.
This mode will be offered as a “Quick Start” option and will not interfere with existing advanced users’ ability to customize their setups.
Describe alternatives you've considered
No response
Additional context
No response