Invoke

What is Invoke

Invoke is a generative media platform designed for creative production and media. It allows users to train and deploy AI models, produce assets, and share automated workflows.

Features of Invoke

• Studio-grade Control: Offers advanced options for precise image manipulation and iteration.
• Invoke Canvas: A workspace with non-destructive layers, targeted editing, and an infinite artboard.
• Layer-based Editing: Enables editing on separate, movable layers for precise control.
• Composite Anything: Allows mixing of pixel paint, reference images, and AI control layers.
• Infinite Canvas: A single, scrollable space for managing multiple ideas, styles, and models.
• Collaboration Tools: Provides invite-only Studio spaces, role-based access, permissions, and shared images for teams.
• AI Model Training: Enables development, refinement, and ownership of custom models based on user IP.
• Base Model Finetuning: Allows creation of custom base models fine-tuned to organizational brand and IP.
• Specialized Model Creation: Facilitates development of specialized models (e.g., LoRA) for specific use cases.
• Base Model Upgrades: Supports building and storing training data for easy updates with new base models.
• Enterprise Deployment: Offers secure, scalable deployment of generative AI tools in isolated environments.
• Verified Security and Compliance: Tested and approved by InfoSec teams, SOC-2 Compliant (Type 2).
• Licensed for Commercial Use: All technology stack components permit commercial use.
• Data Ownership: Users retain full ownership of their data, creations, and models.

Use Cases of Invoke

• 3D Blockout to Concept Render: Generates refined concept renders from rough 3D blockouts.
• Turn 3D Renders into Key Art: Transforms in-game screenshots or 3D renders into 2D key art.
• Sketch to Concept Art Render: Converts sketches into fully rendered concept art pieces.
• Style Consistency: Generates images incorporating game brand guidelines and IP.
• Sketch or Render to Visualization: Transforms architectural sketches or 3D renders into photorealistic works.
• Regional Prompting: Applies different guidance types (compositional structure, prompts, references) to specific image areas.
• IP-Appropriate Assets: Generates assets referencing existing team-created work.

FAQ

• How much does Invoke cost? Invoke offers plans starting at $15/month, with a free open-source Community Edition for local installation.
• Is Invoke cloud-based? Invoke offers both cloud-based products (Indie, Premier, Enterprise) and an open-source Community Edition for local installation.
• What foundational base models does Invoke use? Invoke uses models including SDXL, Stable Diffusion 1.5, Flux.1, Imagen 3 & 4, and ChatGPT-4o. Users can upload their own models, and enterprise customers can have custom models created.
• Is there a free or open-source version of Invoke? Users can trial Invoke for free for 10 days. The Community Edition is available for free download and local installation.
• Can I use my own models with Invoke? Yes, Invoke supports openly licensed SD or SDXL models, Flux.1 models, LoRAs, embeddings, ControlNets, and VAEs. Models intended to copy the likeness of individuals or the style of living artists without consent are prohibited.
• Can I use the images I created in Invoke for commercial use? Yes, all images created with Invoke are 100% owned by the user and can be used commercially.
• Does Invoke share my generated images with other users or train models based on the images I generate? No, Invoke does not share generated images with other users or use them to train other models. Creations remain exclusively yours.
• Can I upload custom models from other sites? Yes, Invoke supports openly licensed SD or SDXL models, Flux.1 models, LoRAs, embeddings, ControlNets, and VAEs. Models intended to copy the likeness of individuals or the style of living artists without consent are prohibited.
• Does Invoke use my data to train its AI models? Invoke does not own proprietary AI models and does not use customer data to train or improve other customers' AI models.