README

Statamic ChatGPT is a Statamic addon that allows you to generate content from ChatGPT from within the Bard editor from a prompt.

Features

Some headline features:

• Generate a complete article from a single prompt
• Generate shorter single paragraphs for when you need help with a portion of an article
• Supports GPT5, GPT5-Mini, GPT5-Nano, GPT4o, GPT4o-Mini, ChatGPT 3.5-turbo, and ChatGPT 4
• Will support future versions when they become available
• Add site-wide context to your prompts to get more tailored responses (for example, configure ChatGPT to think it is an expert on writing SEO-optimised articles about gardening, if that's your thing)
• Configure the maximum number of tokens to be used with each prompt in order to manage your spend
• Manage your API key and various options from within the dashboard

Settings

Icon added to Bard editor

Modal allows you to generate a full article or just a paragraph

Fully formatted article is added to Bard, ready for you to use or edit

Also generate a single paragraph

How to Install

You can search for this addon in the Tools > Addons section of the Statamic control panel and click install, or run the following command from your project root:

composer require bitdigitalcodes/statamic-chatgpt

You will need to add your API key before the addon will work.

How to Use

1. Once installed you should see a new 'ChatGPT' menu option on the left of the control panel. Click it.
2. Enter your API key (https://platform.openai.com/signup if you don't have one already).
3. Select the model to use. Some of the later models may need to make a payment to get access.
4. Select the max tokens, or use defaults.
5. Give ChatGPT a bit of context about your site and the text you want it to write.
6. Save the config.
7. You'll now get a ChatGPT icon on any Bard editor. Click this for access to all features.

Troubleshooting

I'm getting an error mentioning my API key.

You probably haven't set up your API key properly in the control panel. Head to https://platform.openai.com/signup, get an API key, and carefully copy it across. You may also be using an API key that doesn't support the model you want to use, in which case try selecting a different model.

You may need to make a payment before some of the models will work. OpenAI removed the need to do this but there have been some user-land reports of still needing to do it.

Ensure your API key is valid for the models you want to use and you have sufficient credit.

After about 2 minutes I get a message about the API timing out.

ChatGPT takes time to generate a response. We wait 2 minutes before giving up, this is usually plenty of time to get the response but if the API is under extreme load or is down then you may need to try again. If it persists check OpenAI status: https://status.openai.com/

I'm getting an [Object object] error pop-up when using.

Please let us know what steps you took for this to happen. We handle all errors gracefully, and any that are introduced by the API should be displayed to you. If you're getting this error then it's likely something we haven't accounted for. Please raise an issue on Github and we'll look into it.