Recraft: Recraft V3
recraft-ai/recraft-v3
Access Recraft V3 from Recraft using the Puter.js AI API.
Add to your appModel Card
Recraft V3 is a text-to-image model from Recraft, a design-focused AI company founded by Anna Veronika Dorogush, the creator of the CatBoost machine learning library.
It's built around accurate text rendering inside generated images, an area where most image models of its era struggled beyond a few legible characters. It also supports brand style customization through reference images and control over the position and size of text elements within a design.
At launch in late 2024, Recraft V3 took first place on the Artificial Analysis Text-to-Image Leaderboard hosted on Hugging Face, with an ELO score of 1172, ahead of FLUX 1.1 Pro (1143) and Midjourney v6.1 (1093), and held that position for several months.
This is the raster output tier; a separate SVG variant of the same model generates vector graphics directly. It suits general-purpose image generation, marketing visuals, and product imagery.
Cost Per Image $0.04
per generation
Configuration output
resolution
Release Date N/A
Add Recraft V3 to your app
Use Recraft V3 in your app with the Puter.js AI API, no API keys or setup required.
// npm install @heyputer/puter.js
import { puter } from '@heyputer/puter.js';
puter.ai.txt2img("A serene mountain landscape at sunset", {
model: "recraft-ai/recraft-v3"
}).then(image => {
document.body.appendChild(image);
});
<html>
<body>
<script src="https://js.puter.com/v2/"></script>
<script>
puter.ai.txt2img("A serene mountain landscape at sunset", {
model: "recraft-ai/recraft-v3"
}).then(image => {
document.body.appendChild(image);
});
</script>
</body>
</html>
More AI Models From Recraft
Recraft V3 SVG
Recraft V3 SVG is the vector-output variant of Recraft's third-generation model, generating scalable vector graphics (SVG) directly from a text prompt instead of a raster bitmap. It inherits the same design-oriented training as the raster version, including accurate text rendering and brand style customization via reference images. Output is native SVG made of paths and shapes rather than a bitmap that's traced afterward, so it stays sharp at any size and opens directly in vector tools like Illustrator or Figma. Recraft is one of the few widely available text-to-image APIs that outputs vector art directly rather than rasterizing and then tracing it, part of why the underlying V3 model reached the top of the Artificial Analysis Text-to-Image Leaderboard on Hugging Face in late 2024. This tier suits logos, icons, and other vector illustration work where a scalable, editable file matters more than photographic detail.
ImageRecraft V4 Pro
Recraft V4 Pro is the high-end raster tier of Recraft's fourth-generation image model, released in February 2026 as a follow-up to V3. Recraft describes V4 as built around design taste, tuned on composition, lighting, color relationships, and material realism rather than generic photorealism. Compared to V3, Recraft reports improved faces and anatomy, more reliable proportions and expressions, and sharper text rendering for uses like signage and packaging. The Pro tier generates at higher resolution, up to 2048x2048 pixels, aimed at print, product, and marketing work where detail needs to hold up under close inspection. The standard (non-Pro) V4 tier generates faster at lower resolution for social and web use; both tiers share the same underlying aesthetic model, differing mainly in resolution and fidelity. It suits production image generation for professional and commercial output.
ImageRecraft V4 Pro SVG
Recraft V4 Pro SVG is the vector-output counterpart to Recraft V4 Pro, generating native SVG files rather than raster images. It carries V4's design-tuned aesthetics (composition, lighting, color, and typography) into vector output, and Recraft describes it as producing finer paths and more geometric precision than the standard (non-Pro) V4 SVG tier. Output is real, editable vector graphics with structured layers, not a rasterized image converted afterward, so files open directly in tools like Illustrator, Figma, or Sketch. Native vector generation, across both the V3 and V4 generations, is one of Recraft's distinguishing features; few widely available text-to-image APIs output SVG directly at all. This tier fits large-scale brand assets, signage, detailed print illustration, and icon or design-system work where a clean, scalable vector file is the deliverable rather than a bitmap.
Frequently Asked Questions
You can access Recraft V3 by Recraft through Puter.js AI API. Include the library in your web app or Node.js project and start making calls with just a few lines of JavaScript — no backend and no configuration required.
Recraft V3 is free to integrate using the Puter.js AI API. With the User-Pays Model, you can add AI to your app for $0, since users cover their own AI usage through their Puter account.
| Price | |
|---|---|
| Per image | $0.04 |
Yes — the Recraft V3 API works with any JavaScript framework, Node.js, or plain HTML through Puter.js. Just include the library and start building. See the documentation for more details.
Add Recraft V3 to your app for free
Developers can integrate Recraft V3 for free using the Puter.js AI API.
With the User-Pays Model, each user covers their own AI usage instead of the developer.