Trends

.ai

Guias

os melhores modelos de IA de código aberto

Como instalar o SD.Next

A heavily modernized fork of the original WebUI that runs Stable Diffusion, SDXL, Flux, Qwen, Chroma and dozens of other model families on NVIDIA, AMD, Intel and Apple hardware.

Geração de imagens

Windows

Mac

Linux

Licença: AGPL-3.0

Site oficial

GitHub

O que é o SD.Next?

SD.Next (formerly automatic) is a fork of the Stable Diffusion WebUI by vladmandic that has grown into a different program. It supports the widest range of hardware of any UI (CUDA, ROCm, DirectML, Intel Arc, OpenVINO, Apple MPS) and the widest range of models through the Diffusers library, with built-in model downloading, a modern interface, control (ControlNet and friends), video and a detailer for faces and hands. It keeps the tab layout people know from A1111 while updating everything underneath.

Choose SD.Next if you have an AMD or Intel GPU, want many model families in one place without building node graphs, or want a fast moving project that adds new models quickly. The number of options can be overwhelming at first; Fooocus or Invoke are simpler.

Sem instalação: rode modelos de IA na nuvem

BitVector Prism é a forma mais fácil de começar: escolha um modelo, digite um prompt e gere em um app web limpo, sem configurar nada. O BitVector também está disponível no Discord e na web (SpyGlass).

Experimentar o BitVector

O que você precisa

Windows, Linux or macOS with Python 3.10 to 3.12 and Git. NVIDIA cards with 6 GB or more VRAM are the smoothest; AMD cards work on Windows (ZLUDA or DirectML) and Linux (ROCm); Intel Arc and Apple Silicon are supported. 16 GB RAM recommended. Models are downloaded on demand, so plan tens of gigabytes of disk space.

Como obter e instalar o SD.Next

Install Python (3.10 to 3.12) and Git. On Windows, tick Add Python to PATH during setup.
Clone the GitHub repository (git clone https://github.com/vladmandic/sdnext) into a folder without spaces.
Run webui.bat (Windows) or webui.sh (Linux and Mac). The script detects your GPU and installs the right PyTorch build; add --use-rocm, --use-directml, --use-zluda or --use-openvino if it guesses wrong.
Open http://127.0.0.1:7860. Use the Models tab > Reference to download a starter model (for example SDXL or Flux) straight from Hugging Face, or copy your own files into models/Stable-diffusion and models/Lora.
In Settings, pick the backend (Diffusers is the modern default) and the UI theme; then restart the server from the UI.

Baixe apenas pelos links oficiais desta página. Instaladores de outros lugares podem estar desatualizados ou adulterados. O ModelTrends.ai não hospeda arquivos.

Sem instalação: rode modelos de IA na nuvem

O PirateDiffusion é exclusivo do Telegram e feito para profissionais: milhares de modelos, LoRAs e workflows por comandos de chat, com geração ilimitada a preço fixo.

Experimentar o PirateDiffusion

Como usar o SD.Next

Select a model in the dropdown at the top. SD.Next detects the model type and adjusts the available settings.
In the Text tab, write your prompt and negative prompt, choose the size and the sampler, and press Generate.
Use the Networks panel (the button right of the prompt) to add LoRAs and embeddings by clicking them, and add the trigger words from the model page.
The Control tab combines img2img, inpaint, ControlNet, IP-Adapter and processors in one place; load a picture, choose a control type and generate.
Turn on the Detailer in the sidebar to fix faces and hands automatically after each generation, and use the Postprocess tab for upscaling.

Dicas e problemas comuns

Read the console log when something fails; SD.Next prints exactly which dependency or model file is missing.
Flux and other large models can be loaded with on-the-fly quantization (Settings > Quantization) to fit 8 to 12 GB cards.
Use git pull (or the --upgrade flag) regularly; the project updates often and new model support comes through updates.
The documentation site has a dedicated page for each GPU vendor with the exact flags to use.

Modelos que você pode rodar no SD.Next

SD.Next runs Stable Diffusion 1.5, SDXL, Flux, Qwen Image, Chroma, Hunyuan and many other families listed in this catalog, plus their LoRAs. It can download reference models by itself.

Modelos Flux

Modelos Stable Diffusion XL 1.0

Modelos Stable Diffusion 1.5

Modelos Qwen

Modelos Chroma

Modelos Hunyuan

Sem instalação: rode modelos de IA na nuvem

Sem GPU, sem downloads, sem Python. Esses serviços rodam os mesmos modelos de código aberto em seus próprios servidores, então você começa a gerar pelo navegador, Discord ou Telegram em um minuto.

Experimentar o BitVector

Experimentar o PirateDiffusion

Links

Estes links levam aos sites dos próprios criadores. O ModelTrends.ai não é afiliado a eles e não hospeda software.

Outros apps locais

Como instalar o ComfyUI

The node-based powerhouse for Stable Diffusion, Flux, Wan video and almost every new open source model. Free, open source, for Windows, Mac and Linux.