POST /v1/analyze request requires a mode parameter. The mode controls which sub-scores the report returns, which research questions get asked, and how the AI scoring weights each factor. The overall score comes back the same shape regardless of mode — it’s the composition underneath that changes.
Choose the mode that matches the decision your user is making.
The four modes
off_grid
For buyers asking “Can I live here independent of utilities?”
rural_residential
For buyers asking “Can I build a primary residence here without surprises?”
recreational
For buyers asking “Will this be a good hunting / camping / weekend property?”
investment
For buyers asking “Is this a good buy from a returns perspective?”
Score scale
All sub-scores and the overall score are integers0–100. The API also returns a confidence value per sub-score:
Treat
low confidence sub-scores as directional — surface them in your UI but don’t make irreversible decisions on them.
Mixing modes in batch requests
POST /v1/batch accepts up to 50 items, and each item declares its own mode. So a single batch can return one off-grid report and one investment report on the same parcel — useful when comparing perspectives. See batch analysis.
Picking a mode
If you’re building a single-purpose product (off-grid marketplace, hunting-land MLS, etc.), pick one mode and stick to it. If you’re building a generalist tool, expose mode selection to your end-users — they know what decision they’re making better than your code does. When in doubt,rural_residential is the safest default — its sub-scores cover the majority of what most buyers care about regardless of intent.