Website: https://dreamtonics.com/synthesizerv/
Came out more recently than [vocaloid], actively developed. Very high quality of sound. Recent updates support combination [concatenative-synthesis] + machine learning/neural network approach. There is a free version available that even includes some "lite" versions of soundbanks, meaning that they can only synthesize one pitch. Even the full version of the software + a full version of one of the voicebanks is still less expensive than [vocaloid] alternatives, however.
There is support for Windows, Mac, and Linux. As I mentioned, the software is reguarly updated with new releases coming out all the time. Just recently in the past several months, a new release came out with many new features.
Synthesizer V has a web demo that has the full editing UI and vocal synthesis engine running in the browser via WebAssembly: https://synthesizerv.com/web/
Downloading + decompiling the Wasm blob reveals some information about the tool. Including strings for different internal SynthV functions and modules, it also shows that the [juce] framework was used as well as Emscripten. It's interesting that many very modern WebAssembly features were used including multi-threading and bulk memory operations which I didn't even know were available yet.
[sleepwalking] is the person who founded Dreamtonics, the company that produces Synthesizer V. Check his page for more information.