Release 1.0.0.2

I've just added Dall-E plugin to use Generations, Edits and Variations endpoints and an example to use them.

UPDATE 1.0.0.1

NOW YOU CAN HIDE YOUR API KEY WITH THE NEW ACTION, "MAKE PROXY REQUEST”

-This update includes a new action, "Make Proxy Request," for all plugins.

-With this action, you can make an OpenAI API call without revealing your secret API key by using a proxy site to make the call on behalf of the client.

-Instructions for setting this up are available in the plugin documentation.

If you have any questions or need help setting this up, please feel free to ask here.

Thanks guys for the kind words.

Thanks for the suggestion faber80, I’ll take a look at that.

For the first update I’ll address the API key security by adding the action “MakeProxyRequest”, so users can enter a custom proxy URL , so it makes the request at a proxy server that retrieves the secret API key from a location outside the root, makes the OpenAI API call on the client's behalf, and finally sends back the response to C3 app.I’ll include a php example to help on this.

Hi, You can use this url:

editor.construct.net/?safe-mode

And uninstall the problematic plugin in the addon manager

Edit:Sorry, maybe I misread and I thought you missed a plugin file so it crashed. So probably the above doesn’t work.

Release 1.0.0.4

OpenAI Chat Vision and OpenAI TTS were added to the suite.

Now it includes:

1-OpenAI Chat:

1. This plugin allows Construct 3 to interact with the OpenAI chat completion endpoint, enabling real-time chat functionality.
2. Perfect for chatbot development, conversational AI, and virtual assistants

2-OpenAI Chat Vision:

1. Enhance your Construct 3 applications with the power of AI vision. This plugin integrates with OpenAI’s vision models, allowing your application to understand and interpret images in a human-like manner.
2. Perfect for chatbot development with vision capability.

3-OpenAI Dall-E:

1. Bring the power of Dall-E to your Construct 3 projects. These plugins allows your project to generate(Dall-E 2 and Dall-E 3), edit and make image variations(Dall-E 2 only) from textual descriptions, opening up a world of creative possibilities.

4-OpenAI Whisper:

1. Whisper is a speech recognition system for making requests to the OpenAI speech to text, transcriptions and translations endpoints.
2. Transcribe audio into whatever language the audio is in.
3. Translate and transcribe the audio into English.

5-OpenAI TTS:

1. The Text-to-Speech (TTS) plugin allows your Construct 3 project to convert written text into spoken words. This can be used to create more immersive experiences, provide accessibility features, and more.

6-OpenAI Chat Completions:

Deprecated

7-OpenAI Completions:

Deprecated

Release 1.0.0.3:

• Ability to stream responses on the OpenAI_Chat plugin only, for regular and proxy requests.

• New condition called 'On stream end' that is triggered when the OpenAI streamed response is complete.

• New Chat_Example_Extended_Download, which is an extended version of Chat with the ability to download the chat as a .txt file.

• New Chat_Example_Stream, which provides an example of how to use the new 'stream' feature.

• New PHP file to be used with the latest version of the OpenAI_Chat_V1014 plugin to ensure compatibility with the 'stream' feature.

***************************************************************

RELEASE 1.0.0.2

I've just added Dall-E plugin to use Generations, Edits and Variations endpoints and an example to use them.

****************************************************************

RELEASE 1.0.0.1 Now you can hide your API key with the new action, "Make Proxy Request”

-This update includes a new action, "Make Proxy Request," for all plugins.

-With this action, you can make an OpenAI API call without revealing your secret API key by using a proxy site to make the call on behalf of the client.

-Instructions for setting this up are available in the plugin documentation.

If you have any questions or need help setting this up, please feel free to ask here

*****************************************************************

Hello everyone, I have just released the OpenAI Plugin Suite. If you're interested, feel free to ask any questions you may have or suggest a feature.

OpenAI Plugin Suite

Seamlessly integrates OpenAI API GPT and Whisper models with Construct 3

https://www.construct.net/en/game-assets/addons/openai-plugin-suite-3317

• Introducing the OpenAI Plugin Suite - a powerful set of 4 plugins designed to seamlessly integrate the OpenAI API with the Construct 3 event system. With just a few events, you can build intelligent chatbots, create dynamic content, and develop virtual assistants with ease. Moreover, the ability to generate text and language provides this suite with tremendous potential for game developers looking to create dynamic dialogues, storylines, and other game design systems.

OpenAI Construct 3 Plugin Suite includes 4 Plugins:

1. OpenAI_Chat_Completions

• Construct 3 plugin for making requests to the OpenAI chat completion endpoint

• Ideal for single completions without memory of previous prompts

• Utilize GPT-3.5 and GPT-4 models for content creation, text generation, and language translation

• Available models: gpt-4, gpt-4-0314, gpt-4-32k, gpt-4-32k-0314, gpt-3.5-turbo, gpt-3.5-turbo-0301

2. OpenAI_Completions

• Construct 3 plugin for making requests to the OpenAI completion endpoint

• Suitable for single completions without memory of previous prompts

• Ideal for content creation, text generation, and language translation

• Available models: text-davinci-003, text-davinci-002, text-curie-001, text-babbage-001, text-ada-001, davinci, curie, babbage, ada

3. OpenAI_Chat

• Construct 3 plugin for making requests to the OpenAI chat completion endpoint with memory of the entire conversation

• Perfect for chatbot development, conversational AI, and virtual assistants

• Harness the power of GPT-3.5 and GPT-4 models for advanced capabilities

• Available models: gpt-4, gpt-4-0314, gpt-4-32k, gpt-4-32k-0314, gpt-3.5-turbo, gpt-3.5-turbo-0301

4. OpenAI_Whisper

• Construct 3 plugin for making requests to the OpenAI speech to text API, transcriptions and translations endpoints.

• Transcribe audio into whatever language the audio is in.

• Translate and transcribe the audio into english.

• Available models: whisper-1

Disclaimer: Please note that the use of OpenAI Construct 3 Plugin Suite requires an OpenAI API key. Additionally, the API calls are subject to a “per token” payment system, which means that users will be charged for each API call made using the plugins. It is the responsibility of the user to ensure that they have a valid OpenAI API key and are aware of the associated costs before using the plugins.

GPT-4 model is only available for users who have access to it.

Addon documentation:

https://construct3openai.carrd.co/

OpenAI Pricing:

https://openai.com/pricing#language-models

OpenAI documentation:

https://platform.openai.com/docs/introduction

OpenAI models:

https://platform.openai.com/docs/models

Edit:

This template has been removed from the asset store. If you are interested in using OpenAI in your Construct 3 projects, please consider using the OpenAI Plugin Suite instead.

https://www.construct.net/en/game-assets/addons/openai-plugin-suite-3317

**************************************************************

Hi guys, I've just released "OpenAI API Construct 3 template" in the asset store!

Template

https://www.construct.net/en/game-assets/game-templates/openai-api-construct-template-3172

OpenAI API Construct 3 template

The purpose of this template is to demonstrate how the following features can be integrated into Construct 3, enabling you to develop your own applications and games using them. Each feature has its own distinct layout and event sheet, providing you with an understanding of how it works and the ability to easily copy or modify it in your own C3 projects

Introducing the OpenAI C3 template - a powerful and easy-to-use template for integrating the OpenAI API into your C3 projects. With this template, you can easily add AI-powered functionality to your apps/games, including:

● Completions feature for creating completions for any prompt to provide a wide range of possible completions, including but not limited to:

• natural language responses
• dialogue continuation
• text summarization
• code generation
• question answering

● Create image feature for generating images based on a given prompt

● Outpainting feature to generate extended images based on a given prompt. This can be used to add additional elements, details, or context to an existing image

● Create image variation feature for creating variations of existing images

● Create image edit feature to erase certain parts of an existing image and generate new elements based on a given prompt

● Seamless integration with C3

The OpenAI C3 template is designed to integrate seamlessly with C3 and utilizes the C3 event system for the majority of its functionality. Even though certain features such as sending images to OpenAI API requires the use of pure JavaScript, the template comes well-commented and easy to understand, so even those parts will pose no problem for users with a basic understanding of JavaScript

Please note that in order to use the OpenAI C3 template, you will need to provide your own OpenAI API key. This key can be obtained by signing up for an OpenAI account, which is free and easy to do. Once you have your API key, simply enter it into the “APIKEY” Global constant in the template and you’ll be ready to start using all of the powerful AI-powered features it provides

If you have any questions about it, feel free to reach out.

Hi, I'm developing an Android/iOS game that requires microphone permission.

Basically on iOS it shows two prompts on the first launch and shows always one prompt on subsequent sessions ignoring the user permission decisions given on previous sessions.

From a user experience point of view it's very annoying to give mic permission every single time the game is launched.

Is anyone experiencing this?And can anyone point me to a possible solution?

Thanks

On Android all works well:

-I request the mic permission

-The prompt shows up

-The user gives permission

-I play the game and I close it

-The next time I launch the game the permission doesn't show up because the user already accepted it on a previous session.

On the other hand, on iOS(Cordova):

-I request the mic permission

-On the first launch it shows one prompt

-The user gives permission

-It show a second prompt

-The user gives permission again

-I play the game and I close it

-The next time and every subsequent session I play the game it shows a prompt again but just one time.

Circum is just a simple game I put together to test the current C3 APK performance(post r247 and the frame scheduling fix), and test Google services, Leaderboards, Achievements and AdMob, and check the Google Console workflow(that thing is a maze!)

Bigger and better games are coming if all those features work smoothly.

Circum

play.google.com/store/apps/details