The QRP Labs kit ordering page says 5W. In fact I believe they are fine for at least 10W and this is noted on the LPF Kit page
10W is well within the specified ratings of the cores according to the original G-QRP club website article (see for PDF of the article). 10W is also well within the voltage ratings of the NP0/C0G low-loss RF capacitors used in the LPF kits.
It may be possible to push it beyond 10W. Tolerance of SWR mismatch would be reduced (and eventually disappear). Another factor is component tolerances and winding differences resulting in variation in inductance values. If the filter starts producing attenuation at frequencies lower than it should... i.e. at the transmit frequency - then remember that attenuation means reduction in output power. The reduced power has to go *somewhere* and in fact it is likely to be largely dissipated as heat in the LPF components.?
I'm not sure how far beyond 10W would be Ok before generating smoke. Anyone who has these QRP Labs 7-element (3 inductors, 4 capacitors) filters on their tiny 1.5 x 0.5 inches board (37 x 13mm) would intuitively not think it right to put 100W through it. Even if everything measured perfectly I wouldn't want to go anywhere near 100W!
A confession... during PA development of the QRP Labs 5W HF PA kit (Class C) , I did put 30W continuous keyed through a QRP Labs LPF kit... No smoke ;-)
