AIモデル · recraft
Recraft V3
Text-to-image generation with long text support and broad style control.
10 CR / RUN · recraft · 8.4M RUNS
概要
Recraft V3 is a text-to-image model built for users who need more than a basic prompt-to-picture workflow. It supports long text generation inside images and can create visuals across a wide range of styles.
Code-named red_panda, Recraft V3 is described on Replicate as state of the art in image generation, with that claim tied to the Text-to-Image Benchmark by Artificial Analysis. On Replicate, it has also seen extensive real-world usage, with more than 8 million runs.
強み
- 01Generates images from text prompts
- 02Supports long text within images
- 03Works across a wide range of visual styles
- 04Offers control through size and aspect ratio inputs
- 05Widely used on Replicate with millions of runs
プロンプト例
A clean modern poster for a jazz festival, with long readable event text, bold typography, deep blue and gold color palette, editorial style
このプロンプトを試すA cozy coffee brand social ad featuring a steaming cup on a wooden table, handwritten-style headline text, warm lifestyle photography look
このプロンプトを試すA futuristic book cover with neon city lights, dramatic atmosphere, title and subtitle rendered clearly, cinematic sci-fi style
このプロンプトを試すAn illustrated travel postcard of Kyoto in spring, cherry blossoms, detailed landmark scene, decorative long caption text, vintage print style
このプロンプトを試す活用例
Posters with readable copy
Create promotional visuals that combine imagery with longer text content. This is useful when the wording needs to be part of the image, not added later.
Style-led brand concepts
Explore different visual directions from a single prompt by changing style inputs. It can help teams quickly compare looks for campaigns, packaging, or social content.
Editorial and cover art
Generate magazine-style visuals, covers, and feature images with stronger control over composition and text elements. This makes it a practical option for concept development.
Flexible format creation
Use size and aspect ratio controls to adapt outputs for different placements. That makes it easier to create assets for feeds, banners, and other layout needs.
よくある質問
関連モデル
