> ## Documentation Index
> Fetch the complete documentation index at: https://docs.cdn-fly.com/llms.txt
> Use this file to discover all available pages before exploring further.

# FLUX.2 [flex]

> Call the Black Forest Labs FLUX.2 [flex] text-to-image model via the Starrise AI API.

FLUX.2 \[flex] is a highly configurable text-to-image model from Black Forest Labs, available through Starrise AI. It offers fine-grained control over sampling steps and guidance scale, ideal for design professionals. The API uses **synchronous calls** — it directly returns Base64-encoded image data.

## Key Capabilities

* **Text-to-image** — Generate images from text descriptions
* **Step control** — Adjust sampling steps (1–50) to flexibly balance speed and quality
* **Guidance scale** — Fine-tune prompt adherence (1.5–10)
* **High resolution** — supports up to 4MP output resolution
* **Reproducible** — Use `seed` for deterministic generation

## Quick Example

```bash cURL theme={null}
curl https://dayu.cdn-fly.com/v1/images/generations \
  -H "Authorization: Bearer YOUR_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "flux-2-flex",
    "prompt": "A cute baby polar bear",
    "size": "1024x1024",
    "inference_steps": 30,
    "guidance_scale": 7.5
  }'
```

```python Python theme={null}
import requests

response = requests.post(
    "https://dayu.cdn-fly.com/v1/images/generations",
    headers={"Authorization": "Bearer YOUR_API_KEY"},
    json={
        "model": "flux-2-flex",
        "prompt": "A cute baby polar bear",
        "size": "1024x1024",
        "inference_steps": 30,
        "guidance_scale": 7.5
    }
)

result = response.json()
# result["data"][0]["b64_json"] contains the Base64-encoded image
```

## Parameters

| Parameter | Type | Required | Description |
| - | - | - | - |
| `model` | string | Yes | Fixed value: `flux-2-flex` |
| `prompt` | string | Yes | Text description for the image |
| `n` | integer | No | Number of images to generate (1–4), default `1` |
| `size` | string | No | Output resolution, e.g. `1024x1024`, supports up to 4MP |
| `inference_steps` | integer | No | Sampling steps (1–50). Low steps (6–20) for rapid prototyping, high (40–50) for max fidelity |
| `guidance_scale` | number | No | Guidance scale (1.5–10). Higher values produce closer adherence to the prompt |
| `output_format` | string | No | `jpeg` or `png` |
| `seed` | integer | No | Random seed for reproducibility |

## Response Fields

| Field | Description |
| - | - |
| `data[].b64_json` | Base64-encoded image data |
| `created` | Creation timestamp |

> **[API Reference](/en/api-reference/model-api/blackforestlabs/flux-2-flex)**
>
> View the interactive API Playground for FLUX.2 \[flex].


This documentation is built and hosted on [Mintlify](https://mintlify.com), a developer documentation platform.