CouchDB

Open Source, and freely able to install it on any OS you desire (the big 3, anyways) CouchDB was selected for that, it's early-adoption of JSON and its mobile-friendly environment. Also, I have used it off and on in various non-professional projects, and it was really one of the first exposure to databases in my career

Compared to MongoDB, CouchDB's Map-Reduce paradigm poses a steeper learning curve for SQL users. However, CouchDB's master-master replication is an advantage of implementing a load-balanced solution.
Even though, currently, CouchDB has strong community support, as an open source software, there is no guarantee and no dedicated support. Compared to both DynamoDB and CosmosDB, backed by big names (Amazon and Microsoft respectively) behind them, they can certainly provide better support that definitely comes with a price.

MongoDB and CouchDB are both document stores, but their concurrency models and ability to scale are very different. MongoDB cannot replicate / shard over unreliable links and network partitions have been the cause of data loss in the past. MongoDB has an easier query language that is only beginning to be emulated by the Mango query and index system in CouchDB 2.0. Also, MongoDB tends to have better documentation and its communication protocol is a bit more efficient than HTTP.

It has been 5+ years since we chose CouchDB. We looked an MongoDB, Cassandra, and probably some others. At the end of the day, the performance, power potential, and simplicity of CouchDB made it a simple choice for our needs. No one should use just because we did. As I said early, make sure you understand your problems, and find the right solution.

Some random reading that might be useful:
http://www.julianbrowne.com/article/viewer/brewers-cap-theorem
https://www.couchbase.com/nosql-resources/why-nosql\
https://www.infoq.com/articles/cap-twelve-years-later-how-the-rules-have-changed

S3 blew this out of the water, we can get over 30 files a second, almost no failures, auto backed up, don't need our own server, and a much simpler interface with PHP Laravel.

We looked at MongoDB and Firebase. MongoDB gives us the best working db engine with a very intuitive design. However, it does not work as well offline. Firebase was extremely hard to create searching and indexing. Using a third-party to search didn't work for us or at least it was too hard for us to figure out. CouchDB was the best of all worlds. Even though it is not great at any one thing ... it is second place in every category which makes it the best all around. We can get to market faster which is the key driver.

Developing spring boot apps with Couchbase using Spring Data JPA really solves complexity of programming.

I have briefly used MongoDB in other products, and it proved that it had better integration capabilities with Ruby on Rails and node.js software platforms, more than CouchDB. But I never had the chance to actually replace CouchDB with MongoDB in the current product to see what improvements it could have made. But I expect API libraries support would be better.

It stacks up well against Mongo DB. Mongo DB definitely has more marketing and developer and customer mindshare because it is so widely known.

CouchbaseDB is essentially faster with more options but of course with a cost. Cloudant is the hosted option with a better security schema.