Recraft: Recraft V4.1 Vector

Recraft: Recraft V4.1 Vector is an image generation model from Recraft, the vector (SVG) variant of Recraft V4.1. It is tuned for high aesthetics, which means the visual quality of the generated artwork is a deliberate focus of the model rather than an afterthought. If you already know Recraft as a provider of design oriented image models, this is the version aimed specifically at vector output.

The model accepts text and image inputs and produces image output. Its context window is 65,536 tokens, with a maximum output of 58,982. The output format is SVG, so what you get back is vector artwork rather than a raster bitmap. Generation is controlled through image parameters: aspect_ratio accepts 1:1, 4:3, 3:4, 16:9, 9:16 or auto; output_format accepts svg; n controls how many images are produced and ranges from 1 to 6; and input_references ranges from 0 to 1, so you can supply at most one reference image alongside your text prompt.

Those capabilities point to a fairly specific set of use cases. Because the output is SVG, the results scale cleanly to any size, which suits logos, icons, illustrations, diagrams and other assets that need to stay sharp across screen and print. The aspect ratio options cover square, landscape and portrait layouts, so you can target social formats, banners, slides or print proportions directly. Passing a reference image lets you steer style or composition from an existing visual, and generating up to six images in one call is useful when you want to compare directions before committing to one.

You call the model through api14's OpenAI-compatible API using the model id "recraft/recraft-v4.1-vector". Because the endpoint follows the OpenAI format, existing OpenAI client libraries and tooling work without a rewrite. Note that the model lists no supported parameters in the usual text sense; the controls that matter here are the image parameters described above, covering aspect ratio, output format, image count and reference input. Set those according to the asset you need, and keep in mind that the only output format available is SVG.

One practical limit worth planning around is the single reference image: input_references maxes out at 1, so workflows that depend on combining several reference visuals will need a different approach. The output format is fixed to SVG as well, so if you need a raster image such as PNG or JPEG, this model is not the right fit and a raster image model would serve you better. For vector work, though, the combination of text and image input, flexible aspect ratios and multi image generation covers most design asset needs.