How we read the ocean.

We pull live swell, wind and tide data from established global ocean models, the same primary sources most surf forecasts are built on. The difference is what we do with it after.

Generic forecasts give you the open-ocean reading at a grid point offshore. We tune every spot for its own geography. Orientation, shelter, reef vs beach, wind funnelling, tide windows, so the numbers reflect what the wave will actually do when it lands.

Through the day, in line with the underlying model runs. You'll usually see refreshed data every few hours.

The first 3 days are the most reliable. Days 4-7 are a strong steer for planning. Beyond a week, treat it as a signal that something might be on the way, not a promise.

Groundswell is the primary long-period swell that's travelled in from a distant storm, the organised lines that actually shape the surf. Sea is the total significant wave height, combining groundswell with any locally generated wind-sea sitting on top. A 2ft groundswell can still produce a chest-high wave because the sea component fills in the rest. Most forecasters quote that combined figure as their headline 'swell', we show both so you know what kind of size you're getting.

Period is the time between waves and it's the single best clue to wave quality. Short periods (under ~9s) usually mean weak, choppy windswell. Mid periods (10-13s) start to form proper lines. Long periods (14s+) carry serious energy and refract into spots that shorter swells miss entirely.

Each break has a 'window' of directions it likes. A 4ft swell pointing at the wrong angle can be flat at one spot and pumping a mile up the coast. We score every spot against the live swell direction so you can see who's lit up and who's shadowed.

Wind speed alone isn't the whole story, what matters is its angle relative to the beach. Offshore grooms waves, onshore flattens them, side-shore depends on the spot. We resolve wind against each spot's orientation rather than just showing a compass direction.

Spots have tide windows where they work best, some need a pushing mid, some only break on a low, some get fat at high. We blend tide state into the rating per spot, not as a generic 'high/low' badge.

We weigh swell size, period, direction, wind angle, wind strength and tide state, all against that spot's preferences. The rating is what comes out when all those factors agree (or don't). We're not going to walk through the exact weighting, but the inputs aren't a secret.

Because we're rating the wave at the beach, not the swell offshore. A clean 3ft groundswell with the wrong wind and wrong tide isn't a good session, and our rating reflects that even when raw numbers look tempting.

Bodhi is the AI local sitting on top of the forecast. He reads the same data you see, factors in your ability and the boards in your quiver, then tells you where to go, what to ride and when to paddle out, in plain language, the way a veteran at the carpark would.

Yes. The call you get as an improver is different from the one an experienced surfer gets for the same conditions. He won't send a beginner into reef that's over their head, and he won't waste an experienced surfer's time on a session that'll bore them.

South Wales, from the Severn Estuary across to Pembrokeshire, including Porthcawl, Gower, Swansea Bay, Carmarthenshire, the Vale and the west coast breaks.

Honest answer: no forecast is perfect, the ocean has the final word. We keep tuning spot-by-spot based on what actually happens in the water, and we publish accuracy updates as we improve the model.

Tell us. Spot tuning is an ongoing job and ground-truth from surfers who know a break is the best feedback we get, it's how the model gets sharper.