Invoke (InvokeAI) is a free, open source image generation app with a professional, Photoshop-like interface. It has a unified canvas where you paint, inpaint, outpaint and regenerate regions on layers, a model manager that installs models from Hugging Face or Civitai by link, and a workflow editor for node graphs when you want them. The Community Edition is the free local app; the company also sells a hosted version, which is why the app feels more finished than most.

Invoke suits artists, designers and anyone who wants to work on an image rather than just generate one. It is also one of the easiest installs with a proper launcher for Windows, Mac and Linux. Power users who need the latest experimental models or video will still want ComfyUI next to it.

Aucune installation requise – exécutez les modèles IA dans le cloud

BitVector Prism est la façon la plus simple de commencer : choisissez un modèle, tapez un prompt et générez dans une application web épurée, sans rien configurer. BitVector est aussi disponible sur Discord et sur le web (SpyGlass).

Essayer BitVector

Ce qu’il vous faut

Windows 10 or 11, macOS 12 or later (Apple Silicon recommended) or Linux. NVIDIA GPU with 6 GB VRAM (8 GB for SDXL, 12 GB or more for Flux); AMD on Linux with ROCm; Apple Silicon through MPS with 16 GB unified memory or more. About 10 GB for the program plus your models.

Comment obtenir et installer Invoke (InvokeAI)

Download the Invoke Community Edition launcher for your system from the official site (invoke.com) or the GitHub releases.
Run the launcher, choose an install folder on a drive with space, pick your GPU type and let it install. It handles Python and PyTorch for you.
Start Invoke from the launcher. The browser interface opens at http://127.0.0.1:9090.
Open the Model Manager (the cube icon). Use Starter Models to install SD 1.5, SDXL or Flux with one click, or paste a Civitai or Hugging Face link, or scan a folder of your existing checkpoints.
Updates are done in the launcher; it keeps your models and settings.

Téléchargez uniquement depuis les liens officiels de cette page. Les installateurs trouvés ailleurs peuvent être obsolètes ou modifiés. ModelTrends.ai n’héberge aucun fichier.

Aucune installation requise – exécutez les modèles IA dans le cloud

PirateDiffusion est réservé à Telegram et conçu pour les pros : des milliers de modèles, LoRA et workflows pilotés par commandes de chat, avec génération illimitée à prix fixe.

Essayer PirateDiffusion

Comment utiliser Invoke (InvokeAI)

Choose a model in the left panel, write your prompt and negative prompt, set the size and press Invoke to generate to the gallery.
Add a LoRA from the LoRA section in the left panel, adjust its weight and include the trigger words from the model page in your prompt.
Switch to the Canvas to work on an image: select a region with the bounding box, paint a mask and press Invoke to inpaint only that area.
Drag the bounding box outside the image and generate to outpaint; Invoke blends the new area with the old one.
Use Control Layers to add ControlNet guidance (canny, depth, pose) or IP-Adapter reference images, and Regional Guidance to give different prompts to different areas.
Right click any gallery image to recall its prompt and settings, upscale it, or send it to the canvas.

Conseils et problèmes courants

Flux in Invoke needs the quantized versions on cards under 24 GB; the starter models include a suitable one.
If models from another app are not detected, use Scan Folder in the Model Manager and tick in-place so they are not copied.
The Workflows tab gives you ComfyUI-style nodes inside Invoke for repeatable pipelines.
Low on VRAM? Lower the canvas bounding box size and let Invoke upscale the final result instead.

Modèles utilisables dans Invoke (InvokeAI)

Invoke runs Stable Diffusion 1.5, SDXL (including Pony and Illustrious) and Flux checkpoints and LoRAs from this catalog, plus ControlNet and IP-Adapter models. It does not run the video models.

Modèles Flux

Modèles Stable Diffusion XL 1.0

Modèles Stable Diffusion 1.5

Aucune installation requise – exécutez les modèles IA dans le cloud

Pas de GPU, pas de téléchargement, pas de Python. Ces services exécutent les mêmes modèles open source sur leurs propres serveurs : vous générez depuis un navigateur, Discord ou Telegram en une minute.

Essayer BitVector

Essayer PirateDiffusion

Liens

Ces liens mènent aux sites des éditeurs. ModelTrends.ai n’est pas affilié à eux et n’héberge aucun logiciel.

Autres applis locales

Comment installer 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.