Stable Diffusion
AI模型
图像
上一代
August 2022: image generation for everyone, weights and all.
您正在查看一条 AI 模型记录:谁构建了它、它能做什么,以及其背后的硬件与数据。
Stable Diffusion's open release — trained on LAION's web-scraped billions — put text-to-image on consumer GPUs and spawned the fine-tune/LoRA/ControlNet ecosystem, the artist-consent backlash, and the lawsuits that still shape training-data law. SD3.5 (2024) continues the line under Stability's community license.
它如何与整个技术栈相连
本记录的关联链接一览——点击分组可直接跳转。
开发方
训练数据包括
后继产品
来源
记录审阅截至 2026-01-15 — 查看 方法论.