Dollars. Miles. Possibilities.
Bring cash and award options into one search. Keep points, cash, and taxes explicit so your agent can explain the tradeoffs.
FLIGHT API FOR AI AGENTS
A world of flights.
One API.
Ready for agents.
From “where can I go?” to “that’s the one.”
Cash fares, award seats, and flexible trips — built for the way
agents search.
In development. A preview of what’s taking off.
A BETTER WAY TO GET THERE01 / FF
“London in April. Cash or points?”
ONE SEARCH. MORE POSSIBILITIES.EXAMPLE
JFKNew York
12 APR 2027
ONE WAY · ECONOMY
LHRLondon
PAY WITH CASH$485.00
or
USE YOUR POINTS35,000 pts+ $120.00 in taxes & fees
Sample fares · not live availability
DESIGNED FOR THE WHOLE SEARCH
Cash + award flights
Flexible exploration
x402 pay-per-search
LESS GLUE CODE. MORE GOING PLACES.
Flight answers your
agent can work with.
Travel starts with a question. We’re building the layer that turns it into structured options, clear tradeoffs, and a useful next step.
Room for “somewhere warm.”
Search a defined set of destinations and dates. Give an open-ended request a clear scope, then show exactly what was checked.
Context, alongside the fare.
Know when an offer was seen, what a search covered, and what’s missing. An empty result should never mean “we didn’t look.”
FROM INTENT TO ITINERARY
Your agent plans.
FlyFetch finds.
A flight-shopping workflow designed to fit into your agent’s existing tools.
Take a look under the hood-
01
Search with a clear scope
Set the route, travelers, dates, and constraints. Explore bounded alternatives when the plan is still taking shape.
-
02
Compare with the full context
Read structured offers with fare types, taxes, freshness, and known limitations. Retrieve retained results without starting another paid search.
-
03
Refresh. Then hand off.
Recheck the chosen offer and get the best supported booking link or instructions. The traveler completes the purchase with the provider.
BUILT FOR BUILDERS
A familiar API.
A new kind of traveler.
REST resources. Structured responses. Explicit search costs. A small surface your agent can learn and compose.
DRAFT CONTRACT · NOT LIVE YETcurl -X POST \
https://api.flyfetch.dev/v1/searches \
-H "Idempotency-Key: your-unique-request-id" \
-H "Content-Type: application/json" \
-d '{
"fare_types": ["cash", "award"],
"legs": [{
"origins": ["JFK"],
"destinations": ["LHR"],
"departure_date": "2027-04-12"
}],
"travelers": {"adults": 1},
"constraints": {"cabin": "economy", "max_stops": 0},
"display_currency": "USD",
"execution": {"allow_partial_coverage": true}
}'BEFORE YOU BOARD
A few good
questions.
Can I use FlyFetch today?
Not yet. FlyFetch is in development. This site previews the product and proposed API; the examples do not query live fares. Availability, coverage, and pricing will be published before access opens.
Is this for cash flights or award flights?
Both are part of the design. Cash fares and award seats have different pricing and availability, so results will keep those differences explicit. Launch coverage depends on approved data access and will be declared by the API.
Will FlyFetch book or issue tickets?
The initial product focuses on shopping: search, exploration, comparison, and refreshing offers. Booking handoff will send travelers to a supported provider or give clear next-step instructions. Direct booking and ticket servicing are outside the initial scope.
How will agents pay for searches?
The planned model is x402 pay-per-search, with the cost made clear before a search runs. Reading retained results and polling for completion will not trigger a new search charge. Pricing and supported payment networks are still being finalized.
Will there be an MCP server?
MCP tools are on the roadmap. The REST API is the foundation, so agents and applications can use the same underlying search workflow regardless of their tool framework.
NEXT STOP: BETTER FLIGHT ANSWERS