Name: FLUX.2 [klein] 4B Base Image Generation API | api.market
Brand: magicapi
SKU: api-magicapi-optimized-ai-fine-tuning-black-forest-labs-flux-2-klein-4b-base
Availability: InStock
Author: API Market

Shashank Agarwal

# 🖼️ Text-to-Image and Image Editing API (FLUX.2 [klein] 4B Base)

## Overview
FLUX.2 [klein] 4B Base is the un-distilled version of FLUX.2 [klein] from Black Forest Labs. Its developer optimized it for fine-tuning, customization and post-training workflows.

This API runs the base model for image generation. It does not train or fine-tune a model. Send a text `prompt` for text-to-image, or add up to 5 input images in `images` for image-to-image generation. You choose the output size in megapixels, the aspect ratio, the guidance scale and the output format. Each request returns an array of image URLs.

## Features
- **Text-to-image and image-to-image**: a prompt alone, or a prompt plus up to 5 input images (jpeg, png, gif or webp).
- **Output size**: `output_megapixels` of `0.25`, `0.5`, `1` (default), `2` or `4`.
- **Aspect ratios**: `1:1` (default), `16:9`, `9:16`, `3:2`, `2:3`, `4:3`, `3:4`, `5:4`, `4:5`, `21:9`, `9:21`, or `match_input_image` to follow the first input image.
- **Guidance**: classifier-free guidance from 1 to 10 (default 4). Higher values follow the prompt more closely.
- **Go fast**: `go_fast` runs the prediction with additional optimizations.
- **Seed**: set a seed to reproduce a result.
- **Output formats**: `jpg` (default), `png` or `webp`, with `output_quality` from 0 to 100 for jpg and webp.

## Use Cases
1. **Testing the base model**: see how the un-distilled FLUX.2 [klein] 4B responds to your prompts before planning your own fine-tuning or post-training.
2. **Text-to-image generation**: product shots, food photography, illustrations and concept images from a prompt.
3. **Image-to-image**: guide a new image with one or more reference images.
4. **Size and cost control**: pick a smaller `output_megapixels` for drafts and a larger one for final images.

## Visual Examples
### Example Input
Prompt: "A bowl of ramen on a wooden counter, steam rising, food photography"

All other fields were left at their defaults (no input images, `output_megapixels` `1`, `aspect_ratio` `1:1`, `output_format` `jpg`).

### Example Output
![FLUX.2 klein 4B Base output: a bowl of ramen on a wooden counter with steam rising](https://blog.api.market/wp-content/uploads/2026/10/flux-2-klein-4b-base-output-api-market.jpg)

## Input Parameters
| Field | Type | Required | Default | Allowed values | Description |
| --- | --- | --- | --- | --- | --- |
| `prompt` | string | Yes |  |  | Text prompt for image generation. |
| `images` | array of URLs | No | `[]` |  | List of input images for image-to-image generation. Maximum 5 images. Must be jpeg, png, gif, or webp. |
| `aspect_ratio` | string | No | `1:1` | `1:1`, `16:9`, `9:16`, `3:2`, `2:3`, `4:3`, `3:4`, `5:4`, `4:5`, `21:9`, `9:21`, `match_input_image` | Aspect ratio for the generated image. Use 'match_input_image' to match the aspect ratio of the first input image. |
| `output_megapixels` | string | No | `1` | `0.25`, `0.5`, `1`, `2`, `4` | Resolution of the output image in megapixels |
| `guidance` | number | No | `4` | 1 to 10 | Classifier-free guidance scale. Higher values produce images more closely related to the prompt. |
| `go_fast` | boolean | No | `false` |  | Run faster predictions with additional optimizations. |
| `seed` | integer | No | none |  | Random seed. Set for reproducible generation |
| `output_format` | string | No | `jpg` | `webp`, `jpg`, `png` | Format of the output images |
| `output_quality` | integer | No | `95` | 0 to 100 | Quality when saving the output images, from 0 to 100. 100 is best quality, 0 is lowest quality. Not relevant for .png outputs. |
| `disable_safety_checker` | boolean | No | `false` |  | Disable safety checker for generated images. |

## Pricing
- **Pay As You Go (PAYG) pricing**: you only pay for what you use.
- **Price**: $0.012 per API unit.
- **1 API unit = 1 megapixel.** A request is billed for the megapixels of the output image plus the megapixels of every input image in `images`, at the same rate. The total is rounded up to a whole unit per request, with a minimum of 1 unit.
- The output size is set by `output_megapixels`. With no input images:
  - `0.25`: 1 unit (the minimum), $0.012
  - `0.5`: 1 unit (rounded up), $0.012
  - `1` (default): 1 unit, $0.012
  - `2`: 2 units, $0.024
  - `4`: 4 units, $0.048
- **Worked example**: `output_megapixels` `1` with one 1000 x 1000 pixel input image (1 megapixel) is 1 + 1 = 2 units = $0.024. With `output_megapixels` `2` and two such input images it is 2 + 2 = 4 units = $0.048.
- Aspect ratio, guidance, go fast, seed and output format do not change the price.
- **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/optimized-ai-fine-tuning-black-forest-labs-flux-2-klein-4b-base`

### 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/optimized-ai-fine-tuning-black-forest-labs-flux-2-klein-4b-base/predictions" \
  -H "Content-Type: application/json" \
  -H "x-api-market-key: YOUR_API_MARKET_KEY" \
  -d '{
    "version": "2289efa5ebba21f5322ba1b73ac92bb6fec9f34bafc08e0c26f465dac6f8b465",
    "input": {
      "prompt": "A bowl of ramen on a wooden counter, steam rising, food photography"
    }
  }'
```

Response (201 Created):
```json
{
  "id": "ufawqhfynnddngldkgtslldrkq",
  "version": "2289efa5ebba21f5322ba1b73ac92bb6fec9f34bafc08e0c26f465dac6f8b465",
  "status": "starting",
  "created_at": "2026-10-13T10:55:19.482Z"
}
```

#### Check Prediction Status
```bash
curl -X GET "https://prod.api.market/api/v1/magicapi/optimized-ai-fine-tuning-black-forest-labs-flux-2-klein-4b-base/predictions/{id}" \
  -H "x-api-market-key: YOUR_API_MARKET_KEY"
```

Response when the prediction has finished (200 OK):
```json
{
  "id": "ufawqhfynnddngldkgtslldrkq",
  "status": "succeeded",
  "output": [
    "https://blog.api.market/wp-content/uploads/2026/10/flux-2-klein-4b-base-output-api-market.jpg"
  ],
  "error": null,
  "created_at": "2026-10-13T10:55:19.482Z",
  "started_at": "2026-10-13T10:55:20.482Z",
  "completed_at": "2026-10-13T10:55:27.482Z"
}
```

`output` is an array of image URLs.

### 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`, more than 5 input images, or an `output_megapixels` value that is not in the list.
- **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/optimized-ai-fine-tuning-black-forest-labs-flux-2-klein-4b-base";

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: "2289efa5ebba21f5322ba1b73ac92bb6fec9f34bafc08e0c26f465dac6f8b465", 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({
      "prompt": "A bowl of ramen on a wooden counter, steam rising, food photography"
  }).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/optimized-ai-fine-tuning-black-forest-labs-flux-2-klein-4b-base"

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": "2289efa5ebba21f5322ba1b73ac92bb6fec9f34bafc08e0c26f465dac6f8b465", "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(create_and_wait_for_prediction({
    "prompt": "A bowl of ramen on a wooden counter, steam rising, food photography"
}))
```

## FAQ
- **Does this API fine-tune or train a model?**
  - No. It runs the FLUX.2 [klein] 4B Base model for image generation. The developer built this base version for fine-tuning and post-training workflows, which happen outside this API.
- **How many input images can I send?**
  - Up to 5, in jpeg, png, gif or webp format. Each input image adds its megapixels to the bill.
- **How do I keep the aspect ratio of my input image?**
  - Set `aspect_ratio` to `match_input_image`. It follows the first image in `images`.
- **How do I lower the cost of a request?**
  - Use a smaller `output_megapixels` and smaller input images. Requests are billed per megapixel with a minimum of 1 unit.
- **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 or need assistance, you can reach out to us through the embedded chat on the bottom right corner of the page or send an email to hello@magicapi.com.


Run FLUX.2 [klein] 4B Base from Black Forest Labs: text-to-image and image-to-image with up to 5 input images, 0.25 to 4 megapixel output. $0.012 per megapixel.