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

# Recraft V4

> Recraft V4 is a text-to-image model optimized for design-quality outputs, brand assets, and illustrations. Developers should reach for it when they need reliable, composition-aware image generation at a low cost.

## Overview

Recraft V4 is a text-to-image generator from Recraft that ranks at the top of the HF Arena. It focuses on design-aware composition, lighting, and textures, making it suitable for brand assets and illustrations.

On Kyma, it operates as a frontier-open model in the cheap cost tier with medium generation speed. It accepts text prompts and returns images via an OpenAI-compatible endpoint using a single API key. Requests benefit from automatic failover if a serving path degrades, and exact usage costs are returned in the response payload. Prompt caching applies to repeated prefixes, billing them at this model's cached input rate.

The model does not support vision inputs, reasoning, or structured outputs. It is strictly a text-to-image pipeline, and token-based context limits do not apply.

## Specs

| Field              | Value                                              |
| ------------------ | -------------------------------------------------- |
| Model ID           | `recraft-v4`                                       |
| Best for           | Design-quality default, brand assets, illustration |
| Context window     | 0                                                  |
| Input modalities   | Text                                               |
| Output modalities  | Image                                              |
| Tool calling       | Yes                                                |
| Structured outputs | Yes                                                |
| Prompt caching     | Yes                                                |
| Speed              | Medium                                             |
| Cost band          | Cheap                                              |
| Release stage      | Stable                                             |

## Pricing

|        | Per 1M tokens |
| ------ | :-----------: |
| Input  |     \$0.20    |
| Output |     \$0.60    |

## Use this when

* **Brand asset generation** — Create consistent marketing visuals and logos from text descriptions.
* **Editorial illustration creation** — Produce design-aware artwork with controlled lighting and textures.
* **Rapid UI prototyping** — Generate placeholder images for interface mockups and design workflows.
* **Commercial stock imagery** — Output high-quality images for product catalogs and web content.

## Not ideal for

Do not use this model for tasks requiring image understanding, multi-step reasoning, or structured JSON outputs, as it only accepts text prompts and generates images.

## Pick something else when

* You need vector graphics output: use [`recraft-v4-vector`](/models/recraft-v4-vector) or [`recraft-v4-vector-pro`](/models/recraft-v4-vector-pro).
* You need faster image generation: use [`imagen-4-fast`](/models/imagen-4-fast) or [`flux-1.1-ultra`](/models/flux-1.1-ultra).
* You need image-to-image editing: use [`flux-kontext-pro`](/models/flux-kontext-pro) or [`recraft-v4-pro`](/models/recraft-v4-pro).

## Example

```python theme={null}
from openai import OpenAI

client = OpenAI(base_url="https://kymaapi.com/v1", api_key="ky-...")

response = client.chat.completions.create(
    model="recraft-v4",
    messages=[{"role": "user", "content": "..."}],
)
```

## Agent query example

Ask the API which models fit, instead of hardcoding an id:

```bash theme={null}
curl "https://kymaapi.com/v1/models?recommended_for=generate-images&tools=true&quality_tier=frontier-open"
```

## FAQ

**Does Recraft V4 support image inputs or vision tasks?**

No. It only accepts text prompts and outputs images. For vision or multimodal inputs, choose a different model.

**How does prompt caching affect billing on Kyma?**

Repeated prompt prefixes are billed at this model's cached input rate, reducing costs for workflows that reuse system prompts or style guides.

**What happens if the generation endpoint experiences latency?**

Kyma automatically reroutes the request to a healthy serving path, so you do not need to implement manual retry logic.