On one motor or a cluster?
I was really tempted at first to say either my Quest RTF Space Shuttle or my RTF Space Fighter :dark: , but upon weighing them, they aren't even close.
For clusters, it would have to be my FlisKits Diminutive Deuce. (But hey, it uses 2 engines! And I have repaired it a couple of times and repainted it.) With the new paint job and with a 0.5" wide x 0.25" tall nylon washer that I installed as an ejection piston, it weighs just under 20 grams. That weight is without motors.
For a single Micromaxx II, it is a virtual dead heat between my FlisKits Tumble Weed (with adapter), my Quest RTF Saturn V, my ASP Micro Sandhawk, and a couple of scratch-builds: a "reel rocket" made from the reel from an ice fishing pole, and my Bic Stic Payloader. They are all at just about 10 grams in weight without motor. The Payloader is nearly twice the length of the typical Bic Stic rocket; it has a frosted clear payload section made from a Bic Stic Grip pen barrel. I flew it with a tiny 1.5" long cyalume light stick (used for fishing) in the payload compartment at dusk once. It has a somewhat upscaled version of the fin can designed by Art Applewhite. For purposes of this post, I weighed it without a motor but with a light stick.
The Reel Rocket is 1.625" in diameter. The reel is made from a solid piece of Delrin. Other parts of the rocket include a motor mount, an airframe, a nose cone, a small cardstock aerodynamic shroud and a taped-on launch lug. I wanted to make it so that I could completely disassemble the rocket and reassemble the reel in its original form. The nose cone is comprised of the reel's small plastic handle, plugged with a piece of 3/16" dowel. It sits atop a 3" BT-2.5 airframe to improve the stability margin. The conical shroud just slides over the airframe and down to the top of the reel. The reel has stub for the handle molded into one edge on one side; the stub is parallel to the axis of the reel. The plastic handle slips over it and is held on by friction. I constructed the launch lug by bonding a FlisKits micro launch lug to a short section of thin dowel (as a stand-off) and then bonded that to a standard 1/8" lug, and then slit the larger lug down the opposite side. I slip the larger lug over the handle stub and wrap it with a piece of masking tape to secure it.
The hole in the center of the reel is large enough to fit a Micromaxx motor (loosely) but is too small for the BT-2.5 to fit through. To prep the rocket, I detach the reel from its rod mount and pull the handle off of its stub. I wrap the nozzle end of a Micromaxx II motor with a few wraps of 0.25" wide masking tape to form a thrust ring, and then insert it through the center hole of the reel. I haven't decided yet whether it is better to have the handle stub/launch lug pointing aft or pointing forward. The direction is determined by which side of the reel I insert the motor through, and it causes the launch lug to be either above the reel or below it. At any rate, I wrap the forward end of the motor that protrudes out of the center hole with some more masking tape, and then friction-fit the airframe onto it. Then I slide the conical shroud down over the airframe tube until it's base is against the reel. I prepare the nose cone by sliding the reel's handle onto the 1.5" long dowel, and it is held in place via friction. The bottom 0.5" of the dowel is covered by a section of BT-2, which forms the nose cone shoulder. A 12" long piece of 28 lb. Kevlar thread is tied onto the dowel just above the shoulder; at the other end of the Kevlar shock cord, a small rubber band is tied on. I slip the rubber band on around the outside of the reel and then fit the nose cone on. The rocket recovers via nose-blow recovery. I attach the launch lug as previously described.
The Reel Rocket is tied with my MMX-adapted Tumble Weed and my RTF UFO for the most "majestic" boost.

Among these three, though, it is the lowest flier. It shoots off the pad but starts to slow dramatically just after it clears the launch rod, and then it slowly reaches an apogee of about 5 feet, where it hovers for almost a second before tumbling back down. It ejects the nose cone just after it starts to descend, and the recoil helps to induce even more tumbling. Despite the slow lift-off speed, it ascends straight up and it remains pointed straight up all the way to apogee.
The other three rockets are all decent fliers, with the ASP Micro Sandhawk being the performance champ. That rocket, in fact, flies really well and gets appreciable altitude. In its one flight so far, the Bic Stic Payloader exhibited a bit of coning as it coasted to apogee.
MarkII