In recent years, the field of quantum computing has significantly developed in both the improve- ment of hardware as well as the assembly of various software tools and platforms, including cloud access to quantum devices. Unfortunately, many of these resources are rapidly changing and thus lack accessibility and stability for robust algorithm prototyping and deployment. Effectively lever- aging the array of hardware and software resources at a higher level, that can adapt to the rapid development of software and hardware, will allow for further advancement and democratization of quantum technologies to achieve useful computational tasks. As a way to approach this challenge, we present a flexible, high-level framework called algo2qpu that is well-suited for designing and test- ing instances of algorithms for near-term quantum computers on the cloud. Algorithms that employ adaptive protocols for optimizations of algorithm parameters can be grouped under the umbrella of “adaptive hybrid quantum-classical” (AHQC) algorithms. We demonstrate the utility of algo2qpu for near-term algorithm development by applying the framework to implement proof-of-principle instances of two AHQC algorithms that have applications in quantum chemistry and/or quantum machine learning, namely the quantum autoencoder and the variational quantum classifier, using Rigetti Computing’s Forest platform.