字节跳动发布 Seedream 5.0 Pro 图像生成模型，支持文本生成和最多 10 张参考图像引导，提供 1K/2K 分辨率，定价每张 0.045-0.09 美元
多源视角
RT Geek<br>Pi-Agent 教程：10 章把 Agent Loop、工具系统、消息系统、事件驱动、会话管理和上下文工程全部拆了一遍。每章都讲三层：概念是什么、源码怎么写的、为什么这么设计。三种阅读方式：Web 在线版带三栏配图、本地 Markdown 和 PDF 离线版。<br><br>https://github.com/buchidonggu…
查看这条来源Re Try here:<br>http://replicate.com/bytedance/seedream-5-pro
查看这条来源AI 摘要
字节跳动发布 Seedream 5.0 Pro 图像生成模型，支持文本生成和最多 10 张参考图像引导，提供 1K/2K 分辨率，定价每张 0.045-0.09 美元。
推荐理由常规快讯，保留列表
原文
Seedream 5.0 Pro is ByteDance’s flagship image model. It generates high-quality images from a text prompt, and can use up to 10 reference images to guide the result for character consistency, style matching, and multi-image composition.
What it does
- Text to image — describe what you want and get a sharp result at 1K (~2 megapixels) or 2K (~4 megapixels).
- Reference-guided generation — pass up to 10 reference images to keep a character, product, or style consistent across generations, or to blend multiple subjects into one scene.
- Flexible aspect ratios — square, portrait, landscape, and wide formats (1:1, 4:3, 3:4, 16:9, 9:16, 3:2, 2:3, 21:9).
- PNG or JPEG output.
Resolutions
- 1K — around 2 megapixels, good for fast iteration and web use.
- 2K — around 4 megapixels (up to 2048x2048, or wider for cinematic ratios), for higher-detail work.
When you pass reference images and leave the aspect ratio on match_input_image, the output matches the shape of your first reference image at the resolution you pick.
Pricing
You pay per output image, by resolution:
- 1K: $0.045 per image
- 2K: $0.09 per image
Model created 1 week, 5 days ago
讨论
暂无评论。