- Image Generation
Recraft V4.1 Vector is a text-to-vector image generation model from Recraft that converts prompts into fully editable SVG graphics with clean geometry and structured layers. It…
Powered by Recraft
Recraft V3
- Image Generation
Recraft V3 is Recraft’s second-generation text-to-image model, notable for native vector output, long multi-word text rendering, and over 20 style presets tailored for professional design workflows.
Providers Comparison
Speed
Output tokens per second · Higher is better
L
LLM API
S
SiliconFlow (FP8)
G
GMI
C
CoreWeave
N
Novita
D
DeepSeek
P
Parasail (FP8)
D
DeepInfra (FP4)
End-to-End Response Time
Seconds · Lower is better
L
LLM API
S
SiliconFlow (FP8)
G
GMI
C
CoreWeave
N
Novita
D
DeepSeek
P
Parasail (FP8)
D
DeepInfra (FP4)
Price
USD per 1M tokens (blended) · Lower is better
L
LLM API
S
SiliconFlow (FP8)
G
GMI
C
CoreWeave
N
Novita
D
DeepSeek
P
Parasail (FP8)
D
DeepInfra (FP4)
About the model
What is Recraft V3?
Recraft V3 is an image generation model from Recraft that produces ~1K-resolution images from text (and optionally image) inputs with advanced control over style, layout, and color. It is mainly used for professional graphic and brand design, including logos, icons, marketing assets, and scalable SVG vector art that stays crisp at any size. It is also used for compositions that require accurate multi-word text placement, such as posters, ads, and UI mockups, reducing manual post-editing. Recraft V3 builds on Recraft V2 and the earlier “red_panda” research line as Recraft’s second-generation flagship image model.
Model capabilities
5 Core Capabilities
-
Text-to-Image Art
Generates high-quality images from text prompts, optimized for professional graphic design, illustration, and commercial visual content creation.
-
Vector Graphic Design
Creates clean, scalable vector-style artwork suitable for logos, icons, branding materials, and other professional design workflows.
-
Accurate Text Rendering
Produces images with clear, readable textual elements, enabling posters, ads, packaging, and other designs requiring embedded lettering.
-
Prompt-Based Styling
Responds sensitively to nuanced text prompts, enabling control over style, composition, and branding aesthetics for creative professionals.
-
Multilingual Prompts
Accepts prompts written in multiple languages, allowing international users to generate images using their preferred natural language.
Use cases
6 Most Valuable Use Cases
- Branding & Logo Design
- Marketing Visual Creation
- Vector Icon Generation
- Poster & Banner Layouts
- Text-Heavy Ad Graphics
- Image-to-Image Restyling
Transparent pricing
Cost Comparison
LLM API offers the lowest per-image cost and best performance for Recraft V3–class models.
| Provider | Region | Latency | Throughput | Uptime | Input ($/1M) | Output ($/1M) | Context |
|---|---|---|---|---|---|---|---|
| LLM API BEST | Global | ~350ms | ~120 img/min | 99.99% | ~$0.0006/img | ~$0.0000/img | ~16 images/batch |
| Recraft | Global | ~600ms | ~60 img/min | ~99.9% | ~$0.0010/img | ~$0.0000/img | ~8 images/batch |
| OpenAI (Image Models) | Global | ~700ms | ~45 img/min | ~99.9% | ~$0.0120/img | ~$0.0000/img | ~10 images/batch |
| Stability AI | Global | ~800ms | ~40 img/min | ~99.5% | ~$0.0080/img | ~$0.0000/img | ~8 images/batch |
| Replicate | US East | ~900ms | ~30 img/min | ~99.0% | ~$0.0100/img | ~$0.0000/img | ~6 images/batch |
Performance benchmarks
Technical Specifications
| Metric | Recraft V3 | Midjourney V8.1 | Stable Diffusion XL 1.0 |
|---|---|---|---|
| Latency per Image | ~1.5s | ~3s | ~3s |
| Throughput | ~40 img/min | ~20 img/min | ~30 img/min |
| Max Resolution | ~2048x2048 | ~2048x2048 | 1024x1024 |
| Price per Image | ~$0.015 | ~$0.020 | ~$0.004 |
| Supported Formats | PNG, JPG, WEBP | PNG, JPG | PNG, JPG, WEBP |
| Uptime | ~99.5% | ~99.0% | ~99.5% |
30-day usage via LLM API
- 1.8B
- Image prompts processed (30 days)
- 42M
- API requests served (30 days)
- 3.6M
- Unique developers & designers (30 days)
- 99.8%
- Avg API uptime (30 days)
Architecture & Integration
Why Build on LLM.API?
One unified API. Every major model. Built-in reliability, cost control, and observability.
-
Intelligent Model Routing
Dynamically route each request to the optimal model across providers based on latency, cost, and quality—without changing your integration or client code.
One endpoint, any model -
Cost-Aware Execution
Enforce per-project and per-request budgets with automatic model selection and pricing visibility, so you control spend while your teams keep shipping fast.
Ship without bill shock -
Resilient Fallback Logic
Define provider and model fallbacks that trigger on errors, timeouts, or rate limits, keeping your AI features online even when upstream APIs fail.
Stay up, even when they’re not -
End-to-End Observability
Trace every call across providers with logs, metrics, and structured events, making it easy to debug prompts, tune routing, and monitor SLAs from one place.
See every token, everywhere -
Task-Level Abstractions
Express workloads as high-level tasks—chat, tools, embeddings, moderation—and let LLM.API map them to the right models and providers automatically.
Think tasks, not providers -
High-Throughput Batching
Batch thousands of inferences into a single optimized pipeline, reducing per-call overhead and unlocking bulk processing for evals, indexing, and backfills.
More throughput, less overhead
Decision guide
When to Use — When NOT to Use
Use it if...
- You need to generate high-quality digital illustrations or concept art from text prompts.
- You need stylistically consistent branding assets like logos, icons, and social media visuals.
- Your use case involves quickly iterating on design variations for marketing or product visuals.
- You need to create web or mobile UI mockups with clean, modern visual styles.
- Your use case involves generating flat, vector-like graphics suitable for presentations or infographics.
- You need creative visual exploration for storyboards, game assets, or character concepts.
Avoid if...
- You need a text-based LLM for reasoning, code generation, or natural language understanding.
- Your workload requires precise scientific diagramming where domain accuracy outweighs visual style.
- You need photorealistic medical or safety-critical imagery with guaranteed anatomical correctness.
- Your workload requires dynamic, interactive 3D rendering or CAD-level geometric precision.
- You need on-device, offline inference with strict resource limits and no external API calls.
- Your workload requires generating long-form documents, reports, or structured text outputs.
FAQ
Frequently Asked Questions
-
What is Recraft V3?
Recraft V3 is an image generation model by Recraft focused on high-quality, controllable visual content for design and illustration use cases.
-
What modalities does Recraft V3 support?
Recraft V3 supports text-to-image generation, and may support additional image editing capabilities depending on the connected API features.
-
How is Recraft V3 priced when accessed through LLM.API?
Recraft V3 pricing on LLM.API is usage-based per generated image or compute unit, with exact rates defined in your LLM.API pricing plan.
-
What is the context window or input size for Recraft V3 prompts?
Recraft V3 accepts relatively short textual prompts plus optional configuration parameters, with exact length limits defined by the LLM.API endpoint.
-
How fast is Recraft V3 in terms of latency?
Recraft V3 typically returns images within a few seconds, although latency varies with resolution, configuration, and current LLM.API load.
-
How do I call Recraft V3 through LLM.API?
You call Recraft V3 via the LLM.API HTTP endpoint by specifying the Recraft V3 model name, your API key, and a text prompt payload.
-
What is Recraft V3 best suited for compared to other image models?
Recraft V3 is best suited for clean, stylized, design-oriented images such as graphics, icons, branding, and product visuals.
-
How does Recraft V3 compare to general-purpose multimodal LLMs?
Unlike general multimodal LLMs, Recraft V3 specializes in image synthesis and style control rather than text understanding or long-form reasoning.
-
Does Recraft V3 support batch image generation through LLM.API?
Yes, you can request multiple images per prompt with Recraft V3, subject to per-request limits defined by LLM.API.
-
What are the key limitations of Recraft V3?
Recraft V3 may struggle with precise text rendering in images, detailed small objects, and scenarios outside its trained artistic or design domains.
Compare
Competitive Models
- Image Generation
Recraft V4.1 is Recraft’s fourth‑generation image generation model, offering high‑quality raster and vector outputs for creative and commercial design workflows.
- Image Generation
GPT-5 Image Mini is an OpenAI model for lightweight image understanding and generation, optimized for speed and efficiency over maximum fidelity. It is designed for everyday…