An AI briefing on every model's route

Eleven models, eleven routes: which to sail? Signed in, open the Briefing and an AI compares them for you, in plain words.

Comparing the models is the right reflex, and it has a cost. Eleven routes on the map, eleven arrivals, eleven sets of charts: one model sends you inshore, another out to sea, a third arrives an hour earlier but beats into the swell for its last night. Reading all of it, leg by leg, before deciding takes a long evening, and a tired skipper at the chart table reads it less well than a fresh one.

Saily WR now writes that reading for you. Open the Briefing beside the legs of the route shown, or in the route log, and an artificial intelligence compares every model's route for the skipper: where the models agree and where they part, the hard moments, the waves and the current met on the way, head seas and foul streams included, and the route to prefer, with why. It knows which model each leg's current comes from, speaks of the nights and of the rain, and when the boat has reefs set, it names the reefs that matter on the route it recommends, with their time and place.

The briefing is open to every account, free: sign in with Google or Apple and it is there, in the language of the page. It is written only while its tab is open, once per computation, as it comes, so a routing you never look at costs nothing to anyone. Print every model puts it first, on a page of its own, before the route logs; in the apps, a briefing written whole stays with its routing, to read again offline, provided you opened it before leaving the network.

Take a passage from Groix to Ushant with a westerly swell running. The routes may agree on the arrival within the hour, and still differ where it matters: the briefing can point out that the fastest one meets a long swell almost on the bow through its last hours, while another, a little slower, keeps it on the quarter and finds a fair stream at the end. That is the choice a skipper makes before casting off, now laid out in a few paragraphs rather than read off eleven charts.

The briefing is written by an AI from the routes alone: it knows what the models say along each route, not what the sky will really do, nor the sea state off a headland the grids are too coarse to see. It can be wrong, so check it against the routes and the charts before relying on it. A new computation forgets it, and writing it needs the network. Read it as a second opinion at the chart table, the decision staying yours.

Find out more