The Module
Having previously dealt with ChatGPT, Claude and Gemini separately i decided to produce a module which will interface with all 3 services.
Hence 'wappler-all-in-one-ai' was produced.
NOTE 'wappler-all-in-one-ai' now also supports the DeepL AI text translation service. This includes an extended schema which is explained in the below post
This module requires the following API keys to be defined in Workflow settings for each service to be used. Missing keys will be ignored unless the service is used
This has a simple interface.
The AI question, maximum tokens and temperature can be dynamic inputs so this can easily be called from a content page
The required AI service is selected from the dropdown (i will deal with dynamic below)
Note the temperature range prompt changes with service
Once selected a context sensitive "Engine" option will be offered
Pre defined options are currently:
Chat GPT
gpt-3.5-turbo ,gpt-4o-mini , gpt-4o
Gemini
gemini-1.5-flash, gemini-1.5-pro
Claude
claude-3-5-sonnet-20240620, claude-3-haiku-20240307, claude-3-opus-20240229
Don't forget to check 'Output'
Run the api action
The module returns the AI reply (up to token max). Note it also echo's back the settings used
{
multiAI1: {
result: "Wappler.io is a visual development platform designed to help developers create web and mobile applications more efficiently. Here are some key points about Wappler:
1. Low-code/No-code platform: It allows developers to",
service: "claude",
engine: "claude-3-5-sonnet-20240620",
maxTokens: 50,
temperature: 0.5
}
}
Dynamic AI Input
There may be occasions where you wish to call this module dynamically and specify the AI service and engine dynamically.
Select "Dynamic" from the AI Service dropdown. This hides the predefined options and allows you to enter your choice
The AI service should be specified comma separated in the format service, engine i.e. 'gpt,gpt-3.5-turbo'. Acceptable service names are shown in the prompt.
In this example i set the AI service and engine in a set value step and select it from the picker.
Running the action returns the reply and echo's back the settings
{
multiAI1: {
result: "Wappler.io is a low-code development platform that allows users to visually create web and mobile applications without the need for extensive coding knowledge. It provides a drag-and-drop interface for designing and developing applications, as well as built-in templates and components to",
service: "gpt",
engine: "gpt-3.5-turbo",
maxTokens: 50,
temperature: 0.5
}
}
"Dynamic" can also be used to specify and AI engines from chatGPT, GeminI or Claude not currently predefined.
