# recraft

Recraft is a London-based AI startup that develops proprietary text-to-image generative models and the Recraft Studio web workspace, focused on creative workflows with vector graphics support and design control. Founded in 2022 by Anna Veronika Dorogush, it emerged from stealth in 2023 and has released several model generations through 2026.

Recraft is a [generative artificial intelligence](https://www.wikiprompt.org/wiki/generative-ai) program and service developed by the London-based startup Recraft, Inc. The company also offers Recraft Studio, a web-based workspace that lets users create and edit images, vectors, and mockups using various text-to-image models. Like models such as Midjourney and DALL-E, the Recraft model generates digital images from natural language prompts, and is specifically tailored for creative workflows, with features that emphasize brand consistency, text fidelity, and layout control.

Recraft's proprietary models underpin both the commercial API and studio interface, with a focus on design-oriented outputs such as scalable vectors, high-resolution raster images, and controlled stylistic variation. The technology is grounded in [deep learning](https://www.wikiprompt.org/wiki/deep-learning) approaches, and the company describes its later versions as ground-up rebuilds to improve prompt accuracy and design taste.

## History and background

Recraft, Inc. was founded in 2022 by machine learning scientist Anna Veronika Dorogush, known for co-creating the CatBoost machine learning library at Yandex. The company emerged from stealth on May 31, 2023, with a public release of its vector graphics generation capability on Product Hunt. On January 17, 2024, TechCrunch profiled Recraft's foundational model for graphic design, noting its emphasis on addressing copyright and ethical concerns associated with [AI](https://www.wikiprompt.org/wiki/artificial-intelligence)-generated imagery.

On October 28, 2024, TechCrunch reported that Recraft's third major model, V3, had topped a crowdsourced benchmark, surpassing Midjourney and OpenAI's DALL-E in overall image quality. On May 5, 2025, Recraft announced a $30 million Series B funding round led by Accel, reporting more than four million registered users at the time of the announcement. The company's trajectory reflects broader shifts in the [large language model](https://www.wikiprompt.org/wiki/large-language-model) landscape, though Recraft focuses specifically on image generation rather than text output.

## Models

Recraft has developed multiple generations of its text-to-image models since 2022. Each generation reflects improvements in fidelity, controllability, and support for both raster and vector outputs. The models are proprietary and accessible through the Recraft API and the Recraft Studio workspace. They are also hosted as an image generation API on external platforms such as fal, Replicate, and Prodia.

### Recraft V2

Recraft V2 was released in March 2024 and was the company's first model trained from scratch. It contained roughly 20 billion parameters and introduced native vector image generation, brand-color conditioning, and improved stylistic consistency for icons and illustrations.

### Recraft V3

Recraft V3 was released in October 2024 and achieved first place on the Artificial Analysis benchmark (hosted on Hugging Face). The model introduced advances in photorealism, improved rendering of multi-word text, and increased responsiveness to detailed descriptive prompts. It also added a parameter allowing users to adjust stylistic intensity within generated images.

### Recraft V4

Recraft V4 was released in February 2026. According to the company, V4 is a ground-up rebuild aimed at improving prompt accuracy and output quality for design workflows, with emphasis on design taste and art-directed results. V4 comes in two versions: V4 for faster iteration and V4 Pro for higher-resolution, print-ready assets (described as 1-megapixel and 4-megapixel output, respectively, with vector variants for each).

## Features and capabilities

Recraft's models support controlled stylistic variation through preset or reference-based guidance and can maintain coherent line, color, or layout across multiple outputs. Key features include vectorization, which converts images into native vector formats; style reference, which guides attributes such as color palette or visual tone; and style mixing, which combines multiple stylistic inputs in a single generation. Inpainting editing allows localized image modification while preserving surrounding content.

Recraft models are designed to interpret detailed natural-language descriptions with attention to composition, style, and text placement. According to third-party reviews and benchmarks, the architecture is similar to diffusion models such as Stable Diffusion and the [transformer](https://www.wikiprompt.org/wiki/transformer)-based approaches used elsewhere in [artificial-intelligence](https://www.wikiprompt.org/wiki/artificial-intelligence). The technical details have not been publicly disclosed.

## Recraft Studio and business model

Recraft Studio is a web-based workspace with an infinite canvas that provides access to generation and editing tools, including raster and vector generation with styles, mockups, vectorization, image upscaling, and natural-language editing. The workspace also provides access to external models such as GPT-4o, Imagen, and Flux.

Recraft operates on a freemium business model: the studio offers a free tier with limited daily credits and paid subscriptions for additional features, while the API uses a credit-based system with unit purchases. The company has primarily developed its own proprietary image models, positioning itself within the broader [generative-ai](https://www.wikiprompt.org/wiki/generative-ai) space alongside contenders like Midjourney and DALL-E.

---
Source: https://www.wikiprompt.org/wiki/recraft
License: CC BY-SA 4.0 (https://creativecommons.org/licenses/by-sa/4.0/)
Last updated: 2026-09-07T21:26:53.699305+00:00