Do I need a GPU or cloud server?

No. Yuga runs on standard shared PHP hosting (cPanel). No GPU, no Python, no VPS required. A $3/month hosting plan is enough.

Is my data private?

Yes. All your training data stays on your server. Nothing is sent to OpenAI, Anthropic, or any external AI service unless you explicitly configure an optional LLM backend.

Can I train it on my own content?

Yes. Train on websites, PDFs, DOCX, CSV, plain text, JSON, or HTML. The crawler discovers and learns from every linked page automatically.

Can I cancel anytime?

Yes. Cancel or downgrade at any time. No contracts, no penalties. Your data is not deleted immediately — you get a grace period.

What counts as an API call?

Every request to a chat, brain_chat, think, reason, agent_run, generate, or yugagen_chat endpoint counts as one API call. Training, status, and session management calls are free.

Can I have multiple AI models?

Yes. Each plan allows a different number of named model instances. The free plan includes 1 model; Enterprise allows unlimited models.