How many Api`s needs AI-Scientist for full Funktion

[fatihaltiok]

I have set three Api form Anthropic Openai and Gemini need more?

[Maleesha-Dinujaya]

Hi @fatihaltiok 👋,

The number of APIs required for AI-Scientist's full functionality can vary based on your intended use cases. However, typically, it integrates with a few key APIs:

Model APIs (LLMs or Custom Models) – For running the AI models (like OpenAI, Hugging Face, etc.).
Data Source APIs – To fetch datasets or real-time data (e.g., OpenML, Kaggle, or any specific domain APIs).
Storage APIs – To store intermediate outputs, logs, or models (could use AWS S3, Google Cloud Storage, etc.).
Visualization/Analytics APIs – For generating reports or dashboards (like Plotly, Grafana, etc.).

For minimal functionality, integration with 1-2 APIs (model + data source) may suffice. But for full automation and advanced workflows, around 4-5 well-configured APIs covering these areas can maximize AI-Scientist's potential.