Exam generation flow

Using the Examplary AI exam generation flow, you can let users of your application use AI to generate questions on a specific subject or a set of source materials.

Presets

Presets allow you to pre-configure certain options for the embed session, so that when your users open the embed, they see a tailored experience. None of these fields are required — if you do not provide any presets, the user will see the default options when they open the embed.

Key	Type	Description
`name`	string	The name of the exam. Plain text.
`subject`	string	The subject of the exam. Plain text.
`studentLevel`	string	The difficulty level of the exam. Expects a student level ID from `/student-levels`.
`intendedDuration`	integer	The intended duration of the exam in minutes (in 15 minute increments).
`taxonomyId`	string	The ID of the taxonomy to use for the exam, from `/taxonomies`.
`language`	string	The language code for the exam (e.g. "en" for English). Defaults to the user's locale.
`sourceMaterialIds`	string[]	An array of source material IDs to pre-select for the exam.
`allowedGenerationQuestionTypes`	string[]	An array of question type IDs to limit the question generation options to.
`context`	string	Additional context or instructions for the exam generation. Plain text.

Example

1. Create an Examplary user for your user

To make sure we can save personal preferences and source materials uploaded by the user to their specific account, and can later allow them to reference these personal details, we require creating a user account in your workspace for each of your users.

POST /users

{
  "email": "my-user@example.com",
  "name": "My User"
}

Store the returned user ID in your system to use with any future embed sessions for that user.

2. Create an embed session

Call the Examplary API to create a new embed session. You can configure presets for the exam, as well as theme options.

The actor field should contain the ID of the Examplary user account you created for this customer.

You can either specify a returnUrl or an allowedOrigin, based on how you want to be notified when generation is completed.

POST /embed-sessions

{
  "flow": "generate-exam",
  "actor": "user_423r9j3r0jeddJA...",
  "presets": {
    "subject": "Mathematics"
  },
  "theme": {
    "primaryColor": "#4f46e5",
    "locale": "en"
  },
  "metadata": {
    "myUniqueIdentifier": "abc1234"
  },
  "returnUrl": "https://app.example.com",
  "allowedOrigin": "https://app.example.com"
}

This returns a response that looks like this:

{
  "id": "embed_session_55S843D7HfNfs9RY48PoTprXnRcz2Vw8Crst64UYrBnz...",
  "status": "pending",
  "embedUrl": "https://app.examplary.ai/embeds/55S843D7HfNf...",
  "flow": "generate-exam",
  "actor": "user_423r9j3r0jeddJA...",
  "enabledResponseModes": ["return_url", "post_message"],
  "createdAt": "2025-12-09T16:52:52.120Z",
  "expiresAt": "2025-12-16T16:52:52.120Z",
  "presets": {
    "subject": "Mathematics"
  },
  "outputs": {},
  "theme": {
    "primaryColor": "#4f46e5",
    "locale": "en"
  },
  "metadata": {
    "myUniqueIdentifier": "abc1234"
  }
}

3. Lead the user to the embed URL

You can either redirect the user directly to the URL, or display it in an iframe.

The latter might be better for the user experience, especially when displayed as a modal. This also allows you to listen to status updates in real time, through the postMessage() API.

const embedUrl = "https://app.examplary.ai/embeds/55S843D7HfNf...";
 
iframe.src = embedUrl;
 
window.addEventListener("message", (event) => {
  // Make sure the message is coming from a trusted origin
  if (event.origin !== new URL(embedUrl).origin) return;
 
  // Handle the message
  const { type, status, outputs } = event.data;
  if (type === "examplary:embed-status-update") {
    if (status === "started") {
      // Fired as question generation progresses — outputs.questionGenerationStatus
      // reflects the current generation step (e.g. "generating_outline",
      // "generating_questions"). Useful for showing a progress indicator.
      console.log("Generation in progress:", outputs.questionGenerationStatus);
    } else if (status === "completed") {
      const examId = outputs.examId;
      console.log("Exam generation completed! Exam ID:", examId);
    } else if (status === "cancelled") {
      console.log("Exam generation was cancelled by the user.");
    }
  }
});

If you're using redirects, you can expect a redirect to your specified returnUrl with one of two query string parameters values:

On success: ?status=completed&examId=exam_dj8948hf98hf43
On failure: ?status=cancelled

If neither of these options work for you, you may also poll the Examplary API for status updates:

GET /embed-sessions/{id}

{
  "id": "embed_session_55S843D7HfNfs9RY48PoTprXnRcz2...",
  "flow": "generate-exam",
  "status": "completed",
  "outputs": {
    "examId": "exam_dj8948hf98hf43",
    "questionGenerationStatus": "completed"
  }
  // ...
}

4. Get exam contents

Once the question generation has completed, you can use the API to retrieve your generated questions.

In Examplary's normal format:

GET /exams/{examId}

Or as a QTI package:

POST /exams/{examId}/export/qti3-zip

5. Cleanup

The embed session will expire automatically after 7 days, but because it gives some limited access to your account, you might want to remove it manually:

DELETE /embed-sessions/{embedSessionId}

You can also delete the generated exam once you don't need it anymore:

DELETE /exams/{examId}