search_hotels
Compare hotel cash and points rates side by side. See cents-per-point value and deal scores to recommend the best value for any location and dates.
Instructions
Search for hotels by location and dates with cash AND points pricing.
Returns hotels with side-by-side cash vs points rates, cents-per-point (CPP) valuation,
and deal scores so you can recommend the best value. This is Gondola's unique advantage —
no other travel search shows both cash and points rates together.
Args:
location: City name, address, or area to search (e.g. "Tokyo", "Manhattan, New York", "near LAX airport").
checkin: Check-in date in YYYY-MM-DD format (e.g. "2026-04-15").
checkout: Check-out date in YYYY-MM-DD format (e.g. "2026-04-20").
num_adults: Number of adult guests. Defaults to 2.
chain_name: Optional hotel chain to filter by (e.g. "marriott", "hilton", "hyatt", "ihg").
Case-insensitive substring match against each result's chain. If nothing matches, the
unfiltered results are returned with an explicit note so you don't keep retrying.
loyalty_programs: Optional list of the user's loyalty programs (e.g. ["hilton_honors", "marriott_bonvoy"]).
When provided, results include personalized earnings and tier benefits like 5th night free.
loyalty_points: Optional dict of program name to points balance (e.g. {"hilton_honors": 250000}).
When provided, results indicate whether the user can afford each hotel with points.
limit: Max number of hotels to return (default 20). The response notes how many more exist
and how to narrow; raise this only when the user explicitly wants a longer list.
Returns:
Formatted list of hotels with cash rates, points rates, CPP valuation, and deal recommendations.
Input Schema
| Name | Required | Description | Default |
|---|---|---|---|
| limit | No | ||
| checkin | Yes | ||
| checkout | Yes | ||
| location | Yes | ||
| chain_name | No | ||
| num_adults | No | ||
| loyalty_points | No | ||
| loyalty_programs | No |
Output Schema
| Name | Required | Description | Default |
|---|---|---|---|
| result | Yes |