Skip to main content

​
Text question processing

Process a text-based quiz question and get the answer.

​
Endpoint

POST /process_question

​
Request examples

curl -X POST http://localhost:3000/process_question \
  -H "X-API-Key: AIzaSyDxxxxxxxxxxxxxxxxxxxxxxxxxxxx" \
  -H "Content-Type: application/json" \
  -d '{
    "question": "What is the capital of France?"
  }'

​
Response

{
  "answers": [
    "Paris"
  ]
}

​
Image question processing

Process a screenshot or image containing a quiz question.

​
Endpoint

POST /process_question

​
Request examples

curl -X POST http://localhost:3000/process_question \
  -H "X-API-Key: AIzaSyDxxxxxxxxxxxxxxxxxxxxxxxxxxxx" \
  -F "[email protected]"

​
Response

{
  "answers": [
    "Option B",
    "The answer is 42"
  ]
}
Supported image formats: PNG, JPEG. Maximum file size: 5MB (server.js:79)

​
Screen monitoring

Automatically detect if a screenshot contains a quiz question and get the answer.

​
Endpoint

POST /monitor_screen

​
How it works

  1. Upload a screenshot
  2. Screen Answerer detects if it contains a quiz question (server.js:349)
  3. If detected, it automatically processes and returns the answer
  4. If not detected, returns detected: false

​
Request examples

curl -X POST http://localhost:3000/monitor_screen \
  -H "X-API-Key: AIzaSyDxxxxxxxxxxxxxxxxxxxxxxxxxxxx" \
  -F "[email protected]"

​
Response (question detected)

{
  "detected": true,
  "answers": [
    "Option C",
    "Machine Learning"
  ]
}

​
Response (no question detected)

{
  "detected": false,
  "message": "No quiz question detected in the image"
}

​
Custom model selection

Use the /process_question_with_key endpoint to specify a custom Gemini model.

​
Endpoint

POST /process_question_with_key

​
Available models

  • gemini-2.0-flash-lite (default) - Faster responses, lower cost
  • gemini-2.0-flash - Balanced speed and accuracy

​
Request examples

curl -X POST http://localhost:3000/process_question_with_key \
  -F "apiKey=AIzaSyDxxxxxxxxxxxxxxxxxxxxxxxxxxxx" \
  -F "model=gemini-2.0-flash" \
  -F "question=What is quantum computing?"

​
Response

{
  "answers": [
    "Quantum computing uses quantum mechanics principles to process information"
  ]
}

​
Complete integration example

Here’s a complete example that handles errors, retries, and different question types. 
class ScreenAnswererClient {
  constructor(apiKey, baseUrl = 'http://localhost:3000') {
    this.apiKey = apiKey;
    this.baseUrl = baseUrl;
  }

  async processQuestion(question) {
    const response = await fetch(`${this.baseUrl}/process_question`, {
      method: 'POST',
      headers: {
        'X-API-Key': this.apiKey,
        'Content-Type': 'application/json'
      },
      body: JSON.stringify({ question })
    });

    if (!response.ok) {
      const error = await response.json();
      throw new Error(`${error.error}: ${error.message}`);
    }

    return await response.json();
  }

  async processImage(imageFile) {
    const formData = new FormData();
    formData.append('image', imageFile);

    const response = await fetch(`${this.baseUrl}/process_question`, {
      method: 'POST',
      headers: {
        'X-API-Key': this.apiKey
      },
      body: formData
    });

    if (!response.ok) {
      const error = await response.json();
      throw new Error(`${error.error}: ${error.message}`);
    }

    return await response.json();
  }

  async monitorScreen(imageFile) {
    const formData = new FormData();
    formData.append('image', imageFile);

    const response = await fetch(`${this.baseUrl}/monitor_screen`, {
      method: 'POST',
      headers: {
        'X-API-Key': this.apiKey
      },
      body: formData
    });

    if (!response.ok) {
      const error = await response.json();
      throw new Error(`${error.error}: ${error.message}`);
    }

    return await response.json();
  }
}

// Usage
const client = new ScreenAnswererClient('AIzaSyDxxxxxxxxxxxxxxxxxxxxxxxxxxxx');

try {
  // Process text question
  const textResult = await client.processQuestion('What is 2+2?');
  console.log('Text answer:', textResult.answers);

  // Process image question
  const imageFile = document.querySelector('input[type="file"]').files[0];
  const imageResult = await client.processImage(imageFile);
  console.log('Image answer:', imageResult.answers);

  // Monitor screen
  const screenResult = await client.monitorScreen(imageFile);
  if (screenResult.detected) {
    console.log('Quiz detected:', screenResult.answers);
  }
} catch (error) {
  console.error('Error:', error.message);
}

​
Rate limiting considerations

Rate limits
  • Per-client: 1 request per 5 seconds (server.js:87)
  • Global: 50 requests per minute (server.js:106)
  • Per-IP: 100 requests per 15 minutes (server.js:48-49)
Implement delays between requests to avoid hitting rate limits.
const delay = (ms) => new Promise(resolve => setTimeout(resolve, ms));

// Wait 5 seconds between requests
await client.processQuestion('Question 1');
await delay(5000);
await client.processQuestion('Question 2');

Build docs developers (and LLMs) love

Get started for freeTalk to us