DeepAI

Go to deepai.org and click Sign Up in the top right corner. You can sign up with email or a Google account. After signing in, go to deepai.org/dashboard — your API key is displayed immediately on the dashboard page under 'API Key'. Click the copy icon to copy it to your clipboard. DeepAI's free tier allows a certain number of requests per month (verify the current limit at deepai.org/dashboard — it was 5,000 at last check but may have changed). For production apps that will be used by many users simultaneously, review the paid plan options at deepai.org/pricing. Keep the dashboard tab open so you can paste the key into Bubble in the next step.

In your Bubble editor, click the Plugins tab in the left sidebar. Click 'Add plugins', search for 'API Connector', and install the plugin by Bubble (it's free). Once installed, click 'Add another API' in the API Connector section. Name the API group 'DeepAI'. This is important: set the Authentication to 'None or self-handled' — NOT 'API Key' or 'Bearer Token'. Using Bubble's built-in auth types would set the header name to 'Authorization', which DeepAI does not use. Under Shared headers, click 'Add a shared header'. Set the key (header name) to api-key exactly — lowercase, with a hyphen, no spaces. For the value, click the blue token icon to add a parameter. Name the parameter deepai_key, check the 'Private' checkbox (critical — this keeps the key on Bubble's servers), and set the default value to your DeepAI API key. Set the Base URL to https://api.deepai.org/api. This is the common base for all DeepAI endpoints.

1{
2  "api_name": "DeepAI",
3  "base_url": "https://api.deepai.org/api",
4  "authentication": "None or self-handled",
5  "shared_headers": {
6    "api-key": "<deepai_key: PRIVATE>"
7  }
8}

Within the DeepAI API group, click 'Add a new call'. Name it 'generateImage'. Set the method to POST and the path to /text2img (which combines with the base URL to form https://api.deepai.org/api/text2img). Set the body type to JSON/Body. Add one body parameter: text — this will hold the user's image prompt. Give it a test default value for initialization: something like 'a golden retriever playing in a field'. Now click 'Initialize call'. Bubble will POST to DeepAI's text-to-image endpoint. If your API key is valid and the request succeeds, the response will look like: { "id": "some-id", "output_url": "https://api.deepai.org/job-view-file/..." } Bubble will parse this response and make the output_url field available in your workflows. This is the URL of the generated image — you'll use it as the dynamic source for a Bubble Image element. Important: DeepAI's image URLs are temporary. They expire after a period of time. If your users need to save or reference generated images later, you must download and store them in Bubble's file storage (or AWS S3/Cloudflare R2) immediately after generation — don't rely on the output_url being available indefinitely.

1{
2  "method": "POST",
3  "url": "https://api.deepai.org/api/text2img",
4  "headers": {
5    "api-key": "<private: deepai_key>",
6    "Content-Type": "application/json"
7  },
8  "body": {
9    "text": "<dynamic: image_prompt>"
10  }
11}

Still in the DeepAI API group, click 'Add a new call'. Name it 'summarizeText'. Set the method to POST and the path to /summarization. Set body type to JSON/Body. Add one body parameter: text — the long-form text to summarize. For initialization, provide a paragraph of text as the default value (at least a few sentences — DeepAI's summarization endpoint works best with longer inputs). Initialize the call. The response shape is: { "output": "summarized text here" }. Bubble will parse the output field for use in workflows. On your Bubble page, add a MultiLine Input element for the long text and a Button labeled 'Summarize'. Add a Text element to display the summary. In the button's workflow, add a step: Plugins → DeepAI - summarizeText. Set the text parameter to the MultiLine Input's value. After the API call step, add a 'Set state' action to store the response's output value in a Custom State on the page (create a custom state of type Text named 'summary_result'). Set the Text element's content to this Custom State. You can optionally add a 'Make changes to a Thing' step to save the summary to a database record — useful if users want to reference their summaries later or compare multiple summarization results.

1{
2  "method": "POST",
3  "url": "https://api.deepai.org/api/summarization",
4  "headers": {
5    "api-key": "<private: deepai_key>",
6    "Content-Type": "application/json"
7  },
8  "body": {
9    "text": "<dynamic: long_text>"
10  }
11}

Click 'Add a new call' in the DeepAI group. Name it 'analyzeSentiment'. Set method to POST and path to /sentiment-analysis. Set body type to JSON/Body with one parameter: text. Initialize with a short sentence like 'I love this product, it works great!'. DeepAI's sentiment analysis response is slightly different from the others: { "output": ["Positive"] } — the output is an array, not a string. Bubble will parse this as a list, not a single value. In your workflow, after the API call step, use 'Result of step X's output:first item' to extract the first element of the array ('Positive', 'Negative', or 'Neutral'). This gives you a plain string you can store in a database field or display in a Text element. For a customer feedback use case, add sentiment analysis to your ticket or feedback submission workflow: after creating the Feedback Thing in Bubble, immediately call the sentiment API with the feedback text, then update the newly created Thing's 'sentiment' field with the first item from the output array. This automatically tags each feedback record as it comes in. To display sentiment results with color-coded indicators, use Bubble's Conditional formatting on a Text or Icon element: when the element's data source contains 'Positive', set the color to green; 'Negative' to red; 'Neutral' to gray.

1{
2  "method": "POST",
3  "url": "https://api.deepai.org/api/sentiment-analysis",
4  "headers": {
5    "api-key": "<private: deepai_key>",
6    "Content-Type": "application/json"
7  },
8  "body": {
9    "text": "<dynamic: feedback_text>"
10  }
11}

DeepAI's generated image URLs are temporary — they point to files hosted on DeepAI's servers and will expire after a period of time. For prototyping and testing, this is fine. For production apps where users expect their generated images to persist, you need to save the images immediately after generation. The recommended approach in Bubble: after the generateImage API call step, add a 'Upload file from URL' action using Bubble's File Uploader or a dedicated workflow. Pass the output_url as the source URL, and Bubble will download the image and store it in Bubble's managed file storage, returning a permanent Bubble file URL. Store this permanent URL in the relevant database record instead of DeepAI's temporary URL. For apps that don't need persistent images (e.g., a one-time generation tool), displaying output_url in an Image element immediately is fine — just don't save it for future reference. Also add privacy rules for any data you store: go to Data tab → Privacy → select your image or content data type. Add rules restricting reads to the record's creator. This is especially relevant if users are generating images from private prompts. RapidDev's team has built dozens of AI-powered Bubble apps using DeepAI and similar tools — if you want to discuss the right architecture for your use case, book a free scoping call at rapidevelopers.com/contact.