NEX provides a suite of controllable AI image generation tools that utilize IP-consistent models trained on securely stored team data, enabling precise customization of images through multimodal inputs. This technology addresses the inefficiencies in creative workflows by allowing teams to rapidly generate high-quality images tailored to specific project requirements, significantly reducing model development costs and turnaround times.
Funding
Funding not disclosed
Founders
Product
Problem
Creative teams face inefficiencies in generating high-quality, customized images for various projects, often incurring significant model development costs and long turnaround times. Existing AI image generation tools lack the precise control and IP consistency needed to meet specific project requirements.
Solution
NEX provides a suite of controllable AI image generation tools designed for teams and enterprises, leveraging IP-consistent models trained on securely stored team data. The platform enables precise customization of images through multimodal inputs, including text, pose, color, and structural references. NEX's creative design suite includes Artboard for image synthesis, Enhance for image upscaling, and Realtime Canvas for rapid ideation. The platform's block-based control system allows users to combine multiple image controls, achieving higher quality outputs compared to alternatives.
Target Audience
NEX targets creative enterprises in media & entertainment, NIL, gaming, fashion, and other industries, as well as teams seeking to leverage AI in their creative workflows.
Features
- IP-consistent model development, from fine-tuning to complete pretraining, using team data
- Block-based control system enabling the use of multiple image controls together (pose, color, structure, etc.)
- Artboard for image synthesis using multimodal prompts
- Enhance for image upscaling
- Realtime Canvas for rapid ideation with an intuitive interface
- Collaborative workspaces for teams to search, store, and generate images together
- Data privacy: organization-tier workspaces are completely siloed, and data is not used to improve models
- Supports various multimodal inputs, including text, pose, color, and structural references