Name: LLaVA v1.6 Vicuna 13B API: Image Question Answering
Brand: magicapi
SKU: api-magicapi-language-and-vision-assistant-yorickvp-llava-v1-6-vicuna-13b
Availability: InStock
Author: API Market

Shashank Agarwal

# 👁️ Image Question Answering API (yorickvp LLaVA v1.6 Vicuna 13B)

## Overview
LLaVA v1.6 (Large Language and Vision Assistant) is an open vision-language model that answers questions about images. This version uses Vicuna-13B as its language model. This listing is packaged by yorickvp.

Send a text `prompt` and, optionally, an `image`. For a multi-turn chat, pass the earlier messages in `history`. The output is a list of text chunks; join them to get the full answer. Answers are limited by `max_tokens` (default 1024).

## Features
- **Visual question answering**: ask what is in an image, what is unusual about it or what to watch out for.
- **Text-only prompts**: `image` is optional, so you can also send plain text prompts.
- **Chat history**: `history` takes earlier messages, alternating roles and starting with the user. Write `<image>` in a message to attach the image to it.
- **Sampling controls**: `temperature` (default 0.2) and `top_p` (default 1).
- **Length control**: `max_tokens` (default 1024); a word is generally 2 to 3 tokens.
- **Chunked text output**: `output` is a list of text chunks that you join in order.

## Use Cases
1. **Image description**: describe photos, screenshots or product images in your own words.
2. **Visual question answering**: ask about objects, actions or details in a scene.
3. **Image-based assistants**: build chat features where users upload a photo and ask follow-up questions.
4. **Content review support**: summarize what an image shows before a person reviews it.

## Visual Examples
### Example Input
An image and the prompt "What should I take into account when visiting this place?", with `temperature` 0.2, `top_p` 1 and `max_tokens` 1024:

![LLaVA v1.6 Vicuna 13B example input image](https://blog.api.market/wp-content/uploads/2026/10/llava-v1-6-vicuna-13b-input-api-market.jpg)

```json
{
  "image": "https://blog.api.market/wp-content/uploads/2026/10/llava-v1-6-vicuna-13b-input-api-market.jpg",
  "top_p": 1,
  "prompt": "What should I take into account when visiting this place?",
  "max_tokens": 1024,
  "temperature": 0.2
}
```

### Example Output
The output chunks joined into one answer (shortened here; the full answer lists 16 points):

```
When visiting a place like the one shown in the image, which appears to be a serene lake with a dock and surrounded by forested mountains, you should consider the following:

1. **Weather Conditions**: Check the weather forecast before you go. Mountain weather can change rapidly, and you don't want to be caught in a storm.

2. **Safety**: Be aware of any potential hazards, such as slippery docks, uneven terrain, or wildlife.

3. **Water Safety**: If you plan on swimming or boating, make sure you know how to do so safely. Wear a life jacket if you're not a strong swimmer.

[...]

Enjoy your visit and remember to respect the natural beauty and tranquility of the place.
```

## Input Parameters
| Field | Type | Required | Default | Allowed values | Description |
| --- | --- | --- | --- | --- | --- |
| `image` | string (URL) | No | none |  | Input image URL (optional) |
| `prompt` | string | Yes | none |  | Prompt to use for text generation |
| `top_p` | number | No | `1` | `0` to `1` | When decoding text, samples from the top p fraction of most likely tokens; lower it to ignore less likely tokens |
| `temperature` | number | No | `0.2` | min `0` | Adjusts randomness of outputs: 0 is deterministic and values greater than 1 are very random |
| `max_tokens` | integer | No | `1024` | min `0` | Maximum number of tokens to generate. A word is generally 2 to 3 tokens. |
| `history` | array of strings | No | none |  | List of earlier chat messages, alternating roles, starting with user input. Write <image> in a message to attach the image to that message. |

## Pricing
- **Pay As You Go (PAYG) pricing**: you only pay for what you use.
- **Price**: $0.08 per API unit.
- **1 API unit = 1 successful request with up to 73 seconds of processing.** A request that takes longer is billed one more unit for each started block of 73 seconds (for example, 147 seconds of processing is 3 units, $0.24).
- Longer answers (higher `max_tokens`) and longer chats take more processing time and can use more units.
- **Example cost**: the example request above took 34.7 seconds of processing, so it used 1 API unit and cost $0.08.
- **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/language-and-vision-assistant-yorickvp-llava-v1-6-vicuna-13b`

### 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/language-and-vision-assistant-yorickvp-llava-v1-6-vicuna-13b/predictions" \
  -H "Content-Type: application/json" \
  -H "x-api-market-key: YOUR_API_MARKET_KEY" \
  -d '{
    "version": "0603dec596080fa084e26f0ae6d605fc5788ed2b1a0358cd25010619487eae63",
    "input": {
      "image": "https://blog.api.market/wp-content/uploads/2026/10/llava-v1-6-vicuna-13b-input-api-market.jpg",
      "top_p": 1,
      "prompt": "What should I take into account when visiting this place?",
      "max_tokens": 1024,
      "temperature": 0.2
    }
  }'
```

Response (201 Created):
```json
{
  "id": "xcmbss082hrmr0d1d1dbtf8344",
  "version": "0603dec596080fa084e26f0ae6d605fc5788ed2b1a0358cd25010619487eae63",
  "status": "starting",
  "created_at": "2026-10-23T10:41:33.366Z"
}
```

#### Check Prediction Status
```bash
curl -X GET "https://prod.api.market/api/v1/magicapi/language-and-vision-assistant-yorickvp-llava-v1-6-vicuna-13b/predictions/{id}" \
  -H "x-api-market-key: YOUR_API_MARKET_KEY"
```

Response when the prediction has finished (200 OK, `output` shortened to its first 31 of 392 chunks):
```json
{
  "id": "xcmbss082hrmr0d1d1dbtf8344",
  "status": "succeeded",
  "output": [
    "When ",
    "visiting ",
    "a ",
    "place ",
    "like ",
    "the ",
    "one ",
    "shown ",
    "in ",
    "the ",
    "image, ",
    "which ",
    "appears ",
    "to ",
    "be ",
    "a ",
    "serene ",
    "lake ",
    "with ",
    "a ",
    "dock ",
    "and ",
    "surrounded ",
    "by ",
    "forested ",
    "mountains, ",
    "you ",
    "should ",
    "consider ",
    "the ",
    "following:\n"
  ],
  "error": null,
  "created_at": "2026-10-23T10:41:33.366Z",
  "started_at": "2026-10-23T10:42:55.555Z",
  "completed_at": "2026-10-23T10:43:30.243Z"
}
```

`output` is a list of text chunks. Join them in order (for example `output.join("")` in JavaScript or `"".join(output)` in Python) to get the full answer.

### 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 a missing `prompt`, a `top_p` outside 0 to 1 or a negative `max_tokens`.
- **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/language-and-vision-assistant-yorickvp-llava-v1-6-vicuna-13b";

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: "0603dec596080fa084e26f0ae6d605fc5788ed2b1a0358cd25010619487eae63", 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;
    }

    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({
  "image": "https://blog.api.market/wp-content/uploads/2026/10/llava-v1-6-vicuna-13b-input-api-market.jpg",
  "top_p": 1,
  "prompt": "What should I take into account when visiting this place?",
  "max_tokens": 1024,
  "temperature": 0.2
}).then((output) => console.log(output.join("")));
```

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

API_KEY = "YOUR_API_MARKET_KEY"
BASE_URL = "https://prod.api.market/api/v1/magicapi/language-and-vision-assistant-yorickvp-llava-v1-6-vicuna-13b"


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": "0603dec596080fa084e26f0ae6d605fc5788ed2b1a0358cd25010619487eae63", "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 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")


print("".join(create_and_wait_for_prediction({
    "image": "https://blog.api.market/wp-content/uploads/2026/10/llava-v1-6-vicuna-13b-input-api-market.jpg",
    "top_p": 1,
    "prompt": "What should I take into account when visiting this place?",
    "max_tokens": 1024,
    "temperature": 0.2
})))
```

## FAQ
- **Do I have to send an image?**
  - No. Only `prompt` is required. Add an `image` URL to ask about a picture.
- **How do I continue a conversation?**
  - Pass the earlier messages in `history`, alternating user and assistant messages and starting with the user. Write `<image>` in the message that the image belongs to, then put the new question in `prompt`.
- **Why is the output a list of strings?**
  - The answer comes back as text chunks. Join them in order to get the complete answer.
- **How do I get shorter or more varied answers?**
  - Lower `max_tokens` to cap the length (a word is about 2 to 3 tokens). Raise `temperature` above the default 0.2 for more varied wording, or set it to 0 for the most repeatable answers.
- **How is a request billed?**
  - $0.08 per request for up to 73 seconds of processing, plus one unit for each further started block of 73 seconds. The example took 34.7 seconds and cost $0.08.
- **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.


Ask questions about images and get text answers with LLaVA v1.6 Vicuna 13B, an open vision-language assistant with chat history. Pay as you go, no monthly fee.