> ## 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 Vector Pro

> Recraft V4 Vector Pro generates native SVG vector graphics at 4MP resolution, built for developers integrating print-ready design pipelines. Reach for it when your application requires scalable assets that maintain sharp geometry at any physical size.

## Overview

This model outputs native SVG files rather than raster bitmaps, delivering 4MP resolution with higher detail and scale than the standard V4 Vector variant. It processes text-only prompts to produce vector paths optimized for commercial printing and large-format output.

On Kyma, it operates as a premium-tier image generation endpoint with medium inference speed and supports prompt caching. Repeated prompt prefixes are billed at this model's cached input rate, and every request routes through an OpenAI-compatible interface with automatic failover. The endpoint does not support vision inputs, structured outputs, or reasoning chains.

## Specs

| Field              | Value                                                           |
| ------------------ | --------------------------------------------------------------- |
| Model ID           | `recraft-v4-vector-pro`                                         |
| Best for           | Print-ready logos, packaging illustrations, large-scale signage |
| Context window     | 0                                                               |
| Input modalities   | Text                                                            |
| Output modalities  | Image                                                           |
| Tool calling       | Yes                                                             |
| Structured outputs | Yes                                                             |
| Prompt caching     | Yes                                                             |
| Speed              | Medium                                                          |
| Cost band          | Premium                                                         |
| Release stage      | Stable                                                          |

## Pricing

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

## Use this when

* **Print-Ready Logo Generation** — Generates scalable brand marks that maintain crisp edges across physical media.
* **Packaging Illustration Design** — Creates clean vector artwork optimized for commercial print and layout workflows.
* **Large-Scale Signage Assets** — Produces high-detail SVGs that scale to physical dimensions without pixelation.

## Not ideal for

Do not use this model for photorealistic raster generation, image-to-image editing, or tasks requiring multimodal vision input.

## Pick something else when

* You need photorealistic raster images: use [`flux-2-pro`](/models/flux-2-pro) or [`imagen-4-ultra`](/models/imagen-4-ultra).
* You need standard vector generation at lower cost: use [`recraft-v4-vector`](/models/recraft-v4-vector).
* You need image editing or inpainting: use [`flux-kontext-pro`](/models/flux-kontext-pro) or [`recraft-v4`](/models/recraft-v4).

## Example

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

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

response = client.embeddings.create(
    model="recraft-v4-vector-pro",
    input=["first document", "second document"],
)
```

## 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 this model output raster or vector files?**

It outputs native SVG vector files, which scale infinitely without losing quality.

**Can I cache prompts to reduce costs?**

Yes, prompt caching is enabled and bills repeated prompt prefixes at this model's cached input rate.

**Does it accept image inputs for editing?**

No, it only accepts text prompts and does not support vision or multimodal inputs.