Run your AI
Decthings allows you to run artificial intelligence from a variety of devices, including personal computers, phones, servers, websites and microcontrollers.
Decthings provides two ways of running your AI:
When deciding whether to run in the cloud or locally, consider the following:
- Running in the cloud is generally a simpler setup process and requires less configuration and programming.
- Additional costs for compute resources used may be applied when running in the cloud. For more details, see pricing.
- Running your model in the cloud adds an additional latency due to the time it takes for the network request to be sent to and from the servers.
- If the local hardware is insufficient, model evaluation may be slow or not possible at all. In the cloud, you can increase the amount of compute resources to fit your needs.
Run the AI in the cloud using a network request
When running in the cloud, your system makes a network request (HTTPS) to Decthings servers. The network request contains the model input parameters, and the cloud will then respond with the model output parameters.
This network request is easiest performed by using an API client. See the evaluate API method reference for more details.
Run locally
Running locally is currently a difficult process. It is something we are actively working on to improve. Check back here in the future!
