I noticed that most of the current image recognition services do not update their internal algorithms regularly. There have been many advances in the field of Artificial Intelligence (especially deep learning), that has brought about techniques and algorithms which push the state of the art in terms of accuracy and speed.
Being an AI researcher myself, I decided to design an API that allows developers to utilize the best techniques available in academia in a production setting. The image recognition API on RapidAPI: https://rapidapi.com/aiddt-deepnet-aiddt-deepnet-default/api/deepnet-image-recognition
You can use the API by invoking the rest endpoint from inside your app.
There is a free tier and the pricing for the paid tiers is slightly-lower/on-par with similar services being offered by larger corporations. If you need a custom image recognition solution, please do mention so in the comments.
Any feedback would be appreciated!