Name: Granite Vision 3.3 2B API: Visual Document Understanding
Brand: magicapi
SKU: api-magicapi-visual-document-understanding-ibm-granite-granite-vision-3-3-2b
Availability: InStock
Author: API Market

Shashank Agarwal

# 📊 Visual Document Understanding API (IBM Granite Vision 3.3 2B)

## Overview
Granite Vision 3.3 2B is a compact vision-language model from IBM. Its developer describes it as designed for visual document understanding: extracting content from tables, charts, infographics, plots, diagrams and more.

Send one or more image URLs in `images` with a text `prompt`; the model returns generated text that describes or extracts what you asked for. `max_tokens` (default 512) caps the answer length.

## Features
- **Document images**: tables, charts, infographics, plots and diagrams.
- **Several images per request**: `images` takes an array of image URLs. The single `image` field is deprecated and ignored when `images` is used.
- **Free input tokens**: image and prompt tokens are not billed; only output tokens count.
- **Length controls**: `max_tokens` (default 512) and `min_tokens` (default 0).
- **Sampling controls**: `temperature` (default 0.6), `top_p` (default 0.9), `top_k` (default 50), penalties, `stop_sequences` and `seed`.

## Use Cases
1. **Table extraction**: turn a table in a scan or screenshot into Markdown text.
2. **Chart reading**: describe a chart and pull out its values.
3. **Diagram description**: explain technical drawings and labelled diagrams.
4. **Document Q&A**: ask a specific question about an infographic or form.

## Visual Examples
### Example Input
![Example input: a product diagram with a pin description table](https://blog.api.market/wp-content/uploads/2026/10/granite-vision-3-3-2b-input-api-market.jpg)

Prompt: "Describe the image". Settings: `max_tokens` 512, `min_tokens` 0, `temperature` 0.6, `top_p` 0.9, `top_k` 50, `presence_penalty` 0, `frequency_penalty` 0.

### Example Output
The chunks of this run joined into one answer (shortened here; the full answer ran to the `max_tokens` limit of 512):

````markdown
### Image Description

The image consists of a detailed diagram of a car's rearview mirror, along with an adjacent table listing the pin descriptions for the rearview mirror.

#### Rearview Mirror Diagram

1. **Connector A**: This is the main connection point on the rearview mirror, where it interfaces with the vehicle's electrical system.
2. **LED Indicator**: Located above the mirror, this LED light illuminates when the rearview mirror is in use.
3. **Reward Looking Sensor**: This sensor is used to assist the rearview mirror in providing a clear view by detecting the presence of objects behind the vehicle.
4. **Forward Looking Sensor**: This sensor is used to assist in the vehicle's forward vision, ensuring that the rearview mirror can provide a clear view of the area behind the vehicle.
5. **ON/OFF Switch**: Located on the rearview mirror, this switch turns the mirror on and off.

#### Pin Description Table

The table provides a detailed pinout for the rearview mirror, with corresponding descriptions:

| Pin | Description         |
|-----|---------------------|
| 1   | Key power           |
| 2   | OEC(+)              |
| 3   | Key out             |
| 4   | -                   |
| 5   | -                   |
| 6   | Battery             |
| 7   | OEC(-)              |
| 8   | GND                 |
| 9   | Reverse signal     |
| 10  | IGN (+)             |

...
````

## Input Parameters
| Field | Type | Required | Default | Allowed values | Description |
| --- | --- | --- | --- | --- | --- |
| `prompt` | string | No | `""` |  | User prompt to send to the model. |
| `images` | array of URLs | No | none |  | Image inputs for the model, as an array of image URLs. |
| `image` | string (URL) | No | none |  | Deprecated single image input. Use `images` instead. Ignored if `images` is used. |
| `system_prompt` | string | No | none |  | System prompt to send to the model. The chat template provides a good default. |
| `min_tokens` | integer | No | `0` |  | The minimum number of tokens the model should generate as output. |
| `max_tokens` | integer | No | `512` |  | The maximum number of tokens the model should generate as output. |
| `temperature` | number | No | `0.6` |  | The value used to modulate the next token probabilities. |
| `top_p` | number | No | `0.9` |  | A probability threshold for generating the output. If < 1.0, only keep the top tokens with cumulative probability >= top_p (nucleus filtering). |
| `top_k` | integer | No | `50` |  | The number of highest probability tokens to consider for generating the output. If > 0, only keep the top k tokens with highest probability (top-k filtering). |
| `presence_penalty` | number | No | `0` |  | Presence penalty |
| `frequency_penalty` | number | No | `0` |  | Frequency penalty |
| `stop_sequences` | string | No | none |  | A comma-separated list of sequences to stop generation at. For example, `<end>,<stop>` stops generation at the first instance of `<end>` or `<stop>`. |
| `chat_template` | string | No | none |  | A template to format the prompt with. If not provided, the default prompt template will be used. |
| `seed` | integer | No | none |  | Random seed. Leave blank to randomize the seed. |

## Pricing
- **Pay As You Go (PAYG) pricing**: you only pay for what you use.
- **Price**: $0.0032 per API unit.
- **1 API unit = 10,000 weighted tokens**; an output token counts 1, an input token counts 0.
- That is **$0 per million input tokens** (input tokens, including image tokens, are free) and **$0.32 per million output tokens**.
- Each request is rounded up to a whole API unit, so the smallest charge per request is $0.0032.
- **Example cost**: the example request used 2,871 input tokens and 512 output tokens. 2,871 x 0 + 512 = 512 weighted tokens, rounded up to 1 API unit = **$0.0032**.
- **Failed or cancelled predictions are not billed.**
- **GET requests to check prediction status** are free and don't count toward API unit usage.
- **No monthly subscription fee.**

## API Documentation

### Base URL
All API requests go to: `https://prod.api.market/api/v1/magicapi/visual-document-understanding-ibm-granite-granite-vision-3-3-2b`

### Authentication
Send your api.market API key in the `x-api-market-key` header:
```
x-api-market-key: YOUR_API_MARKET_KEY
```

### Asynchronous Processing
**IMPORTANT**: This API is asynchronous. The flow is:
1. Send a POST request to `/predictions`. A successful request returns **201 Created** with a prediction `id` and a `status` of `starting` or `processing`.
2. Poll GET `/predictions/{id}` with that ID.
3. Keep polling while `status` is `starting` or `processing`.
4. Stop when `status` is `succeeded` (read `output`), `failed` or `canceled` (read `error`).

A polling interval of 1 second and an overall timeout of 600 seconds (10 minutes) work well.

### Endpoints

#### Create a Prediction
```bash
curl -X POST "https://prod.api.market/api/v1/magicapi/visual-document-understanding-ibm-granite-granite-vision-3-3-2b/predictions" \
  -H "Content-Type: application/json" \
  -H "x-api-market-key: YOUR_API_MARKET_KEY" \
  -d '{
    "version": "ibm-granite/granite-vision-3.3-2b",
    "input": {
      "top_k": 50,
      "top_p": 0.9,
      "images": [
        "https://blog.api.market/wp-content/uploads/2026/10/granite-vision-3-3-2b-input-api-market.jpg"
      ],
      "prompt": "Describe the image",
      "max_tokens": 512,
      "min_tokens": 0,
      "temperature": 0.6,
      "presence_penalty": 0,
      "frequency_penalty": 0
    }
  }'
```

Response (201 Created):
```json
{
  "id": "7zamzc7vqsrmy0d1d0zsatr99c",
  "version": "ibm-granite/granite-vision-3.3-2b",
  "status": "starting",
  "created_at": "2026-10-23T10:13:06.293Z"
}
```

#### Check Prediction Status
```bash
curl -X GET "https://prod.api.market/api/v1/magicapi/visual-document-understanding-ibm-granite-granite-vision-3-3-2b/predictions/{id}" \
  -H "x-api-market-key: YOUR_API_MARKET_KEY"
```

Response when the prediction has finished (200 OK):
```json
{
  "id": "7zamzc7vqsrmy0d1d0zsatr99c",
  "status": "succeeded",
  "output": [
    "### Image Description\n\nThe image consists of a detailed diagram of a car's rearview mirror, along with an adjacent table listing the pin descriptions for the rearview mirror.\n\n#### Rearview Mirror Diagram\n\n1. **Connector A**: This is the main connection point on the rearview mirror, where it interfaces with the vehicle's electrical system.\n2. **LED Indicator**: Located above the mirror, this LED light illuminates when the rearview mirror is in use.\n3. **Reward Looking Sensor**: This sensor is used to assist the rearview mirror in providing a clear view by detecting the presence of objects behind the vehicle.\n4. **Forward Looking Sensor**: This sensor is used to assist in the vehicle's forward vision, ensuring that the rearview mirror can provide a clear view of the area behind the vehicle.\n5. **ON/OFF Switch**: Located on the rearview mirror, this switch turns the mirror on and off.\n\n#### Pin Description Table\n\nThe table provides a detailed pinout for the rearview mirror, with corresponding descriptions:\n\n| Pin | Description         |\n|-----|---------------------|\n| 1   | Key power           |\n| 2   | OEC(+)              |\n| 3   | Key out             |\n| 4   | -                   |\n| 5   | -                   |\n| 6   | Battery             |\n| 7   | OEC(-)              |\n| 8   | GND                 |\n| 9   | Reverse signal     |\n| 10  | IGN (+)             |\n\n..."
  ],
  "error": null,
  "created_at": "2026-10-23T10:13:06.293Z",
  "started_at": "2026-10-23T10:13:06.311Z",
  "completed_at": "2026-10-23T10:13:12.104Z"
}
```

`output` is an array of text chunks. Join them in order to get the full answer. The example above shows the chunks from this run already joined into one string. The answer is shortened here.

### Status Codes
- **201 Created**: prediction created (POST `/predictions`). This is the normal success response.
- **200 OK**: prediction details returned (GET `/predictions/{id}`).
- **400 Bad Request**: invalid input, for example `images` that is not an array of URLs.
- **401 Unauthorized**: missing or invalid API key.
- **404 Not Found**: no prediction with that ID.
- **429 Too Many Requests**: rate limit exceeded.
- **500 Internal Server Error** or **503 Service Unavailable**: temporary server problem; retry later.

## Rate Limits
All plans allow 10 requests per second. Requests above that return `429 Too Many Requests`; back off before retrying and honour any `Retry-After` header. Polling a prediction with GET is free but still counts toward the per-second limit, so poll every 1 to 5 seconds rather than in a tight loop.

## Error Handling
- Check the HTTP status code of every request and read the response body for the error message.
- A prediction can also end with `status` set to `failed`; the `error` field then explains why. Failed predictions are not billed.
- Retry 429, 500 and 503 responses with exponential backoff. Fix the input before retrying a 400.

## Implementation Guide
### Example Code

#### JavaScript
```javascript
const apiKey = "YOUR_API_MARKET_KEY";
const baseUrl = "https://prod.api.market/api/v1/magicapi/visual-document-understanding-ibm-granite-granite-vision-3-3-2b";

async function createAndWaitForPrediction(input) {
  const response = await fetch(`${baseUrl}/predictions`, {
    method: "POST",
    headers: {
      "Content-Type": "application/json",
      "x-api-market-key": apiKey
    },
    body: JSON.stringify({ version: "ibm-granite/granite-vision-3.3-2b", input })
  });

  if (response.status !== 201) {
    throw new Error(`API error: ${response.status} ${await response.text()}`);
  }

  const prediction = await response.json();
  return pollForResults(prediction.id);
}

const delay = (ms) => new Promise((resolve) => setTimeout(resolve, ms));

async function pollForResults(predictionId) {
  const maxTimeout = 600000;
  const pollInterval = 1000;
  const startTime = Date.now();

  while (Date.now() - startTime < maxTimeout) {
    const response = await fetch(`${baseUrl}/predictions/${predictionId}`, {
      headers: { "x-api-market-key": apiKey }
    });

    if (!response.ok) {
      throw new Error(`API error: ${response.status} ${await response.text()}`);
    }

    const prediction = await response.json();

    if (prediction.status === "succeeded") {
      return prediction.output.join("");
    }

    if (prediction.status === "failed" || prediction.status === "canceled") {
      throw new Error(`Prediction ${prediction.status}: ${prediction.error || "Unknown error"}`);
    }

    await delay(pollInterval);
  }

  throw new Error("Prediction timed out after 10 minutes");
}

createAndWaitForPrediction({
    "top_k": 50,
    "top_p": 0.9,
    "images": [
      "https://blog.api.market/wp-content/uploads/2026/10/granite-vision-3-3-2b-input-api-market.jpg"
    ],
    "prompt": "Describe the image",
    "max_tokens": 512,
    "min_tokens": 0,
    "temperature": 0.6,
    "presence_penalty": 0,
    "frequency_penalty": 0
  }).then((output) => console.log(output));
```

#### Python
```python
import time
import requests

API_KEY = "YOUR_API_MARKET_KEY"
BASE_URL = "https://prod.api.market/api/v1/magicapi/visual-document-understanding-ibm-granite-granite-vision-3-3-2b"


def create_and_wait_for_prediction(input_data):
    headers = {"Content-Type": "application/json", "x-api-market-key": API_KEY}
    response = requests.post(
        f"{BASE_URL}/predictions",
        headers=headers,
        json={"version": "ibm-granite/granite-vision-3.3-2b", "input": input_data},
    )

    if response.status_code != 201:
        raise Exception(f"API error: {response.status_code} {response.text}")

    return poll_for_results(response.json()["id"])


def poll_for_results(prediction_id):
    max_timeout = 600
    poll_interval = 1
    headers = {"x-api-market-key": API_KEY}
    start_time = time.time()

    while time.time() - start_time < max_timeout:
        response = requests.get(f"{BASE_URL}/predictions/{prediction_id}", headers=headers)

        if response.status_code != 200:
            raise Exception(f"API error: {response.status_code} {response.text}")

        prediction = response.json()

        if prediction["status"] == "succeeded":
            return "".join(prediction["output"])

        if prediction["status"] in ("failed", "canceled"):
            raise Exception(f"Prediction {prediction['status']}: {prediction.get('error') or 'Unknown error'}")

        time.sleep(poll_interval)

    raise Exception("Prediction timed out after 10 minutes")


output = create_and_wait_for_prediction({
    "top_k": 50,
    "top_p": 0.9,
    "images": [
        "https://blog.api.market/wp-content/uploads/2026/10/granite-vision-3-3-2b-input-api-market.jpg"
    ],
    "prompt": "Describe the image",
    "max_tokens": 512,
    "min_tokens": 0,
    "temperature": 0.6,
    "presence_penalty": 0,
    "frequency_penalty": 0
})
print(output)
```

## FAQ
- **Why does `output` come back as a list?**
  - The answer is returned as an array of text chunks. Join them in order (`output.join("")` in JavaScript, `"".join(output)` in Python).
- **Should I use `image` or `images`?**
  - Use `images`, an array of image URLs. `image` is deprecated and ignored when `images` is set.
- **Why did the answer stop mid-sentence?**
  - It reached `max_tokens` (default 512). Raise `max_tokens` for longer answers; the example run stopped at this limit.
- **How is a request billed?**
  - By weighted tokens at $0.0032 per 10,000: input tokens are free and output tokens count 1, rounded up to a whole unit per request. Most short requests cost $0.0032. The example read 2,871 input tokens (free) and wrote 512 output tokens.
- **Is the 201 response from POST an error?**
  - No. 201 Created means the prediction was accepted. Use the returned `id` to poll GET `/predictions/{id}` until `status` is `succeeded`, `failed` or `canceled`.
- **What if my prediction fails?**
  - Failed predictions are not billed. Read the `error` field, adjust the input and try again.

## Contact Us
If you have any questions, reach out to us through the chat on the bottom right of the page. You can also email our support team at hello@magicapi.com.


Read tables, charts, infographics and diagrams in images with IBM Granite Vision 3.3 2B, a compact vision-language model. Pay as you go, no monthly fee.