Recraft V4
Recraft’s design-first image model — art-directed raster and editable vector SVG generation with strong typographic accuracy.
Overview
What is Recraft V4
Recraft V4 is Recraft’s proprietary text-to-image and vector-generation model, released in February 2026. It emphasizes “design taste,” producing art-directed raster images and editable SVGs with balanced composition, cohesive color, and refined detail rather than generic stock-like outputs. It supports long prompts up to 10,000 characters and integrated typography.
Running it privately on Venice
On Venice you run Recraft V4 with zero prompt retention under an anonymized privacy tier — your generations leave no personal history. You pay a flat $0.05 per image in credits, with no subscription lock-in, making professional design workflows permissionless and private.
Assessment
Strengths and limitations
- Design-tuned outputs that make deliberate aesthetic decisions about composition, color relationships, and detail hierarchy rather than producing generic stock-like renders.
- Native vector SVG generation with structured layers and clean geometry, enabling direct editing in professional design tools.
- Strong prompt adherence with support for detailed descriptions up to 10,000 characters, including integrated typography treated as part of the layout.
- Available in Pro resolution up to 2048×2048 for print-ready assets and high-fidelity scenes.
- Closed and proprietary: no open weights, so self-hosting and fine-tuning are impossible.
- Not uncensored on Venice; outputs respect Recraft’s content policies.
- Slower per-generation latency than lightweight rivals such as Chroma or Lustify.
- No tool use, multi-turn editing, or web-search capabilities — it is a pure text-to-image model.
Samples
Sample outputs
Generated on Venice with our standard prompt suite — the same prompts we run through every model of this type, so you can judge it like-for-like.
A retro travel poster with the bold headline "VENICE" in large condensed serif type, sunset color palette, clean layout
Photorealistic close-up portrait of a weathered fisherman at golden hour, 85mm lens, shallow depth of field, natural skin texture
A small red cube balanced on top of a large glossy blue sphere, with a green cone to the right, plain light-grey studio background
Cozy watercolor illustration of a hillside village in autumn, warm tones, soft paper texture
Capabilities
What it supports
- Text to image
- Image to image
Specifications
Datasheet
- Maker
- Recraft
- Released
- February 2026
- Modality
- Text-to-image (raster & vector SVG)
- Max resolution
- Up to 2048×2048 (Pro); 1024×1024 (standard)
- Open weights
- No — proprietary
- Aspect ratios
- 1:1, 3:2, 16:9, 21:9, 9:16, 2:3, 3:4, 4:5
- Prompt limit
- 10,000 chars
- Privacy on Venice
- Anonymized — prompts not stored
- Available on Venice since
- Feb 2026
API
Call it from your code
Venice exposes this model through the REST API. Queue a generation with the model id.
curl https://api.venice.ai/api/v1/image/generate \
-H "Authorization: Bearer $VENICE_API_KEY" \
-H "Content-Type: application/json" \
-d '{
"model": "recraft-v4",
"prompt": "A serene mountain lake at dawn, photorealistic"
}' --output image.pngPricing
What it costs on Venice
Flat per-image pricing on Venice: $0.05 per generation.
New Venice accounts include a free daily allowance and 500 welcome credits — no credit card required.
Alternatives
How it compares
| Model | Output type | Strongest at | Open weights | Price (Venice) |
|---|---|---|---|---|
| Recraft V4 | Raster & vector SVG | Design taste & brand systems | No | $0.05 / image |
| Flux 2 Max | Raster | Photoreal detail | No | $0.09 / image |
| GPT Image 2 | Raster | In-image text & reasoning | No | from $0.27 / image |
| Chroma | Raster | Speed & open-source flexibility | Yes | $0.01 / image |
The only Venice-hosted option with native vector SVG and design-tuned aesthetics.
Use cases
What it is good for
- 01Brand identity systems and campaign visuals that require consistent color and composition.
- 02Logo and icon creation via native SVG vector output.
- 03Marketing collateral, packaging, and editorial photography with embedded typography.
- 04Print-ready assets using the higher-resolution Pro variant.
- 05Rapid creative iteration where art-directed taste matters more than photorealism.
Prompting
Getting better results
Describe aesthetic direction explicitly — include color palette, lighting mood, and composition style — to leverage the model’s design-taste tuning.
Place desired in-image text inside quotation marks; V4 renders typography as a structural element of the scene.
Use detailed prompts up to 10,000 characters for precise control over detail hierarchy and material realism.
Version history
Current — design-tuned raster and vector generation.
FAQ
Frequently asked questions
Recraft V4 is a proprietary text-to-image and vector-generation model released by Recraft in February 2026. It is built around “design taste,” making intentional aesthetic decisions about composition, color, and detail to produce art-directed raster images and editable SVG outputs.
On Venice, Recraft V4 costs a flat $0.05 per image generation. Optional upscales are $0.02 for 2× and $0.08 for 4×. There is no subscription required; you pay per generation with credits.
You can try it on Venice using the platform’s free-tier credits. New accounts receive welcome credits that cover a number of generations before any paid top-up is needed.
No. Recraft V4 is closed and proprietary, and its weights are not available for download or self-hosting. If you need an open-weights image model on Venice, Chroma is an alternative.
Yes. Recraft V4 can produce native, fully editable SVG files with structured layers and clean geometry. On Venice, the standard raster generation is available; check the app for vector export options.
Choose Recraft V4 when you need design-tuned aesthetics, brand consistency, or native vector SVG output. Choose Flux 2 Max if you prioritize photorealistic detail and do not need vector or typography-centric composition.
Yes. Venice runs Recraft V4 under an anonymized privacy tier — your prompts are not stored, profiled, or used for training. You generate without building a personal history.
No. Recraft V4 is a pure text-to-image generation model. It does not support tool use, multi-turn image editing, or web search on Venice.
Run Recraft V4 privately
No prompt logging. No data used for training.