Fusion, definitely. Even traditional hot-fusion at that.
The Farnsworth Fusor was an device first built back in the mid-60s. For a while there was great hope for it because it *does* allow cheap, easy fusion. No-one was ever able to reach break-even energy production though, despite getting tantalizingly close. Basically your reactor is two concentric electrodes in a vacuum chamber - a big positively charged spherical shell on the outside, and a much smaller "wiffle ball" sphere in the center - usually just a few loops of wire. Ramp the voltage up to 10,000-20,000 volts, then let in a trickle of ionized deutrium or similar. The ions race inwards down the massive potential well and zoom right through the inner "sphere" to collide in a tightly-focussed point in the center of the spheres with enough energy to fuse. And if they fail to fuse, well then they zoom back up the potential well until they reverse course and try again, all without any extra energy input requirements. The problem is the inner electrode - an ion typically needs many "bounces" before it collides just right and fuses, and with each pass through the inner electrode there's a small chance that it will hit one of the wires and lose all its kinetic energy. Despite years of work in lots of labs nobody was ever able to get those losses low enough to hit break-even, though it does make for a convenient fast neutron and gamma-ray source (deutrium-deutreum fusion being one of the worst nuclear reactions known for # of free neutrons produced per Watt of energy released, far worse than fission)
Dr. Bussard (yes, of ramscoop fame), believed he could achieve net power production by using magnetic fields to steer the ions around the wires of the inner electrode, preventing collision losses and finally allowing net energy production, but at that point the Tokamak-based reactors had pretty much cornered the market on research dollars. He did eventually get some small sporadic NAVY funding for what eventually became known as the Polywell Fusor, but died not too long ago before achieving success. His team is still at it though, and while under a standard NAVY publishing embargo the few sentences in the annual budget report suggest that they may be at the cusp of success, even to the point of having perhaps achieved p-B fusion recently, which is a much more challenging reaction (something like 100x higher input energy IIRC), but releases all energy as high-velocity He4 nuclei meaning minimal neutron or gamma radiation (only due to incidental p-p fusion) of any sort and potentially a simple, high efficiency conversion to electricity without resorting to the crude "steam engines" that all existing nuclear reactors rely on.