Amadeus Points of Interest

Name: Points of Interest API
Brand: Points of Interest
Availability: InStock

✓ Official Vendor SpecMaps GeolocationPlaces Searchoauth23 EndpointsREST

Install Jentic One Beta

Connect the Points of Interest to your agent

Jentic One is a self-hosted execution layer for AI agents. It lets your agent call the Points of Interest, or any other public or private API you need. You set the rules, the agent never sees your credentials, and every call is logged.

Two steps, two machines. Install the instance in a safe environment, then register your agent from wherever it runs.

Step 1: Jentic One Host machine

# On the machine that will host your Jentic One instance:
curl -fsSL https://raw.githubusercontent.com/jentic/jentic-one/main/tools/install.sh | sh

Step 2: Agent machine

# On the machine where your agent runs (keep this separate from the instance):
curl -fsSL https://raw.githubusercontent.com/jentic/jentic-one/main/tools/install.sh | sh
jentic register       # connects your agent to your Jentic One instance

Jentic One is in public beta. The setup above keeps your agent separate from the instance, which is what you want before using real credentials: an agent running as the same OS user as Jentic One can read its stored keys directly. Just evaluating? A single local install is fine to start. See the secure deployment guide for the tiers.

Capabilities

What an agent can do with Points of Interest API.

Search points of interest within a radius of a latitude/longitude pair

Search points of interest inside a bounding box defined by north, south, east, west coordinates

Retrieve full detail for a single point of interest by its poisId

GET STARTED

Start building with Points of Interest API

Explore with Jentic

View OpenAPI Document

For Agents

Find named landmarks, restaurants, and attractions around a coordinate or inside a bounding box, with category and popularity metadata for ranking and filtering.

Use for: Find restaurants within 1km of these coordinates, Search for sights inside a bounding box covering downtown Paris, Retrieve full details for a point of interest by its id, List all nightlife venues within 500m of a hotel

Not supported: Does not handle bookings, ticket prices, or live availability - use for discovering and looking up named points of interest only.

Use Cases

Patterns agents use Points of Interest API for, with concrete tasks.

★ Hotel-area attractions panel

Populate a 'things to do nearby' panel on a hotel detail page using the radius search endpoint. Pass the hotel's coordinates and a small radius (typically 1-2km) and render the returned points of interest grouped by category. The detail endpoint lets users tap into a specific landmark for richer information without hitting the search endpoint again. Implementation is two GET calls and renders in seconds.

Call GET /reference-data/locations/pois with latitude, longitude from a hotel record and radius=1 and return the top 10 sights and restaurants.

City itinerary builder

Build a multi-day itinerary by querying the bounding-box endpoint over neighbourhood-sized areas of a city and ranking the returned points of interest by category and popularity. The bounding-box variant is ideal for clustering - a route planner can divide a city into walkable zones and pull top sights for each. Combine with Tours and Activities to add bookable experiences alongside free landmarks.

Call /reference-data/locations/pois/by-square with north, south, east, west bounding the Marais district and return the top 5 sights ranked by popularity.

Detail page for a specific landmark

Once a user taps a point of interest in a list view, call the detail endpoint with that poisId to retrieve the full information needed for a detail page - categories, tags, geocode, and any descriptive metadata available. This pattern keeps list-view payloads small and fetches richer data only on demand. Useful for mobile travel apps where bandwidth matters.

Given poisId=9CB40CB5D0, call /reference-data/locations/pois/{poisId} and return the name, category, and coordinates.

AI travel agent destination explorer

An AI agent answering 'what should I do in Paris on Sunday' uses Points of Interest to ground its recommendations in real, geocoded landmarks rather than generic suggestions. The agent calls Jentic to discover the radius search, executes it for a destination coordinate filtered by category, and returns a ranked list with names, categories, and walking distances - citable, verifiable recommendations rather than hallucinated attractions.

Use Jentic to search 'find points of interest near a location', execute the radius search for the user's destination, and return the top 10 attractions filtered to category=SIGHTS.

Key Endpoints

3 endpoints — the amadeus points of interest api returns named landmarks and attractions around a geographic location, with category labels (sights, restaurants, shopping, nightlife) and popularity ranks.

METHOD

PATH

DESCRIPTION

GET

/reference-data/locations/pois

Search points of interest within a radius

GET

/reference-data/locations/pois/by-square

Search points of interest inside a bounding box

GET

/reference-data/locations/pois/{poisId}

Retrieve a single point of interest by id

GET

/reference-data/locations/pois

Search points of interest within a radius

GET

/reference-data/locations/pois/by-square

Search points of interest inside a bounding box

GET

/reference-data/locations/pois/{poisId}

Retrieve a single point of interest by id

Why Jentic?

Three things that make agents converge on Jentic-routed access.

Credential isolation

Your Amadeus client_id and client_secret are stored once, encrypted, by your own Jentic One instance, which runs the client_credentials exchange and injects a fresh bearer token at execution time. The secret never enters the agent's prompt, logs, or context.

Intent-based discovery

Agents search Jentic by intent such as 'find points of interest near a location', and Jentic returns the radius-search, bounding-box, and detail-lookup operations with their parameter schemas so the agent calls the right one.

Time to first call

FAQs

Specific to using Points of Interest API through Jentic.

What authentication does the Points of Interest API use?

OAuth 2.0 client credentials. Exchange your Amadeus API key and secret for a bearer access token, then pass it as the Authorization header on every call. Through Jentic, the client secret is stored encrypted in the vault and only a scoped token is ever exposed to the agent.

Can I search by area rather than radius?

Yes. GET /reference-data/locations/pois/by-square accepts north, south, east, and west query parameters defining a bounding box, which is useful when you want to query an irregular zone like a neighbourhood rather than a circular radius.

How do I fetch full detail for a single point of interest?

Call GET /reference-data/locations/pois/{poisId} with the id from a previous search result. The detail response is richer than the list-view representation and is the recommended pattern for tap-to-expand UIs.

What are the rate limits for the Points of Interest API?

Amadeus enforces per-second and per-month transaction quotas that vary by environment. Test (test.api.amadeus.com/v1) has low quotas suitable for development; production limits depend on your Amadeus for Developers contract.

How do I find points of interest near a location through Jentic?

Install with pip install jentic, then search 'find points of interest near a location' to discover the radius search operation. Load its schema and execute with latitude, longitude, and optional radius and category. Jentic returns the parsed results ready to render. Get started with Jentic One, the self-hosted execution layer.

Does the test environment cover all cities?

No, the test environment returns POIs for a few selected cities only. For broader coverage you must upgrade to production via the Amadeus for Developers self-service portal.

Can I limit what my agent is allowed to do with the Points of Interest API?

Yes. Because you run Jentic One yourself, your own rules decide which of the three operations the agent may call: the radius search, the bounding-box search, and the detail lookup by poisId. You can restrict the agent to just the operation it needs, such as searching points of interest within a radius, and grant it nothing more. All three endpoints are read-only, so a scoped agent cannot reach any write operation because this API has none.