AI Model Catalogue

Recraft's latest image generation model, built around design taste. Strong prompt accuracy, art-directed composition, and integrated text rendering. Fast and cost-efficient at standard resolution.

Generate production-ready SVG vector images from text prompts. Recraft V4's design taste applied to vector output — clean geometry, structured layers, and editable paths.

Granite-4.0-H-Small is a 32B parameter long-context instruct model finetuned from Granite-4.0-H-Small-Base using a combination of open source instruction datasets with permissive license and internally collected synthetic datasets.

Advanced 20B open-weight reasoning models to customize for any use case and run anywhere.

FLUX LoRA trained on ancient Mediterranean paintings — Greek vase paintings, Minoan frescoes, Egyptian tomb murals. Flat schematic figuration, bold outlines, ceremonial character.

FLUX LoRA trained on paintings by J.-A.-D. Ingres. Porcelain-smooth surfaces, exquisitely elongated forms, precise linear contours, neoclassical mastery.

FLUX LoRA trained on ancient Roman paintings — Pompeii frescoes, encaustic panel portraits. Earthy terracotta tones, flat modelling, direct frontal gazes.

FLUX LoRA trained on Fayum mummy portraits. Striking directness, dark engaging eyes, gilded details, luminous encaustic wax technique.

FLUX LoRA trained on portrait photographs by Nadar. Intimate psychologically revealing studio portraits, soft directional natural light, minimal props.

FLUX LoRA trained on portraits by François Clouet. Precise detailed rendering of rich costume and jewels, flat even lighting, Valois court style.

FLUX LoRA trained on paintings by John Singer Sargent. Dazzling bravura brushwork, shimmering whites, vivid capturing of personality and social status.

FLUX LoRA trained on paintings by Titian. Luminous warm flesh tones, rich crimson and gold drapery, monumental Venetian Renaissance style.

FLUX LoRA trained on portraits by Anthony van Dyck. Elegant elongated figures in satin and lace, subtly melancholic expressions, Baroque court style.

FLUX LoRA trained on paintings by Diego Velázquez. Psychological penetration, muted silver-grey palettes, loose painterly surfaces, Spanish Baroque realism.

FLUX LoRA trained on paintings by Thomas Gainsborough. Feathery delicate brushwork, luminous silks, pastoral English backgrounds.

FLUX LoRA trained on paintings by Bronzino. Cold enamel-smooth surfaces, austere aristocratic sitters, jewelled costumes, Florentine Mannerist style.

FLUX LoRA trained on portraits by Thomas Lawrence, the leading British portraitist of the Regency era. Fluid brushwork, romantic lighting, aristocratic ease.

Segment floorplan images into walls, doors, windows, and kitchen zones using a deep learning model, then extract structured contours and center lines as JSON for downstream applications.

DA3Metric-Large metric depth estimation (Apache 2.0). Returns 16-bit PNG depth map with metric metadata.

Generate detailed SVG vector graphics from text prompts. Recraft V4 Pro's design taste with more geometric detail and finer paths — clean layers, editable output, and scalable to any size.

Bria Expand expands images beyond their borders in high quality. Resizing the image by generating new pixels to expand to the desired aspect ratio. Trained exclusively on licensed data for safe and risk-free commercial use

State-of-the-art video motion quality, prompt adherence and visual fidelity

Compact Powerhouse for Vector Reranking

Classifies images with ResNet-50

Creative Upscale focuses on enhancing details and refining complex elements in the image. It doesn’t just increase resolution but adds depth by improving textures, fine details, and facial features.

Lightning-fast video generation with portrait support, camera controls, and synchronized audio. Up to 20 seconds at 1080p, 4K at 50 FPS.

Kling 3.0 motion control: transfer motion from a reference video to any character image with improved consistency and quality.

The pro version of Qwen Image 2 from Alibaba's Qwen team. Enhanced text rendering, realism, and semantic adherence for high-quality image generation and editing.

120b open-weight language model from OpenAI

A feed-forward neural network that offers a novel approach to visual geometry reconstruction.

Google's fast image generation model with conversational editing, multi-image fusion, and character consistency