Token AI

We present the impact of GenBlur technology and how it delivers a safe browsing experience, free from any explicit or inappropriate content.

The GenBlur application is scheduled for release in 2026 and will be available completely free of charge for Android users.

GenBlur is coming to mobile devices in 2026.

GenBlur is an advanced application that leverages artificial intelligence models and technologies to block harmful sexual content across all applications on mobile phones, providing a safer and more secure digital experience.

Developed by Assem Sabry, GenBlur will be available soon for Android devices, with a web version planned for a later release.

Users will be able to download GenBlur exclusively through the official website: www.genblur.com

We are pleased to announce the start of development on the Replica model series, specialized in text-to-speech generation, first revealed in early 2024. These models are designed specifically for the customer service and call center sector in Egypt and the Arab region.
Our models excel in mastering various Arabic dialects, delivering clarity, fluency, and natural speech. In line with our vision, Replica aims to become one of the most effective AI-powered solutions for customer service and first-line call center operations.
These models will be available at the beginning of 2026 in collaboration with our partners and call center recruitment companies.

We’re thrilled to officially announce our new collaboration between Token AI and Scale.up Level — marking the beginning of a strategic partnership to develop cutting-edge AI solutions that truly serve user needs.

This partnership aims to create innovative and accessible artificial intelligence solutions that empower users across industries. Both teams share a common belief that AI is for everyone, and together, we’re working to bring the power of AI to every hand — making it inclusive, practical, and impactful.

By combining the creativity and technical strength of Scale.up Level with the advanced models and engineering expertise of Token AI, we’re building a strong foundation for projects that address real-world challenges and deliver meaningful value to users.

Both companies share one clear vision: democratizing AI and ensuring that intelligent technologies are not limited to a few but accessible to all. Together, we aim to redefine how individuals and businesses interact with artificial intelligence.

We would like to extend our sincere thanks to Aldahir Córdova and his dedicated team at Scale.up Level for making this collaboration possible. We’re excited about the upcoming joint projects and milestones that will emerge from this partnership.

Stay tuned for exciting updates as Token AI and Scale.up Level unveil the first wave of projects built under this collaboration. The journey has just begun — and together, we’re shaping the future of accessible AI.

Introducing the Token AI Embedding Model Series

We are proud to announce the launch of the Token AI Embedding Model Series, a new generation of intelligent embedding models developed by Assem Sabry, Founder and AI Engineer at Token AI.

This series marks a major step forward in semantic understanding and language representation. It is built to power AI-driven systems such as intelligent search, retrieval, and recommendation engines.

About the Series

The Token AI Embedding Models are designed using advanced Transformer architectures and optimized for performance, scalability, and efficiency across various environments.

Models include:

TokenAI-Embedding-Small – lightweight and efficient for edge and mobile use.

TokenAI-Embedding-Medium – balanced between speed and accuracy for production systems.

TokenAI-Embedding-Large – highly capable and contextual for enterprise-scale tasks.

TokenAI-Embedding-VisionText (coming soon) – combines text and visual understanding.

Key Features

Transformer-based architecture

Supports LoRA fine-tuning

Suitable for developers and enterprises

Open Source Availability

Several models from the series will be released as open source to empower researchers, developers, and startups to build smarter and more context-aware AI systems.

Token AI is an innovative AI startup, formerly known as Reality AI.
Founded by Assem Sabry in 2022, our company recently rebranded to Token AI to better reflect our mission and vision.

We specialize in:

LLMs (Large Language Models)

AI Agents

AI Automation Systems

Chatbots

…and much more.

Our goal is to push the boundaries of what AI can do and make it more accessible, intelligent, and impactful for everyone.

October 18, 2024
Marked the first launch of our project, the Mark - V1 chatbot. It received somewhat positive feedback as a simple chatbot of its kind. However, on January 8, 2025, it underwent a major transformation, evolving into Mark - V2, A groundbreaking update for the project.

Following this update, Mark received extensive feedback, the majority of which was overwhelmingly positive. Thanks to the massive improvements, Mark became significantly smarter, more adaptable in user interactions, and ultimately more beneficial. As a result, its user base grew by an impressive 682%
compared to Mark - V1.

Unfortunately, Mark - V2 was the final update for our major project, and no further versions will be released. Moving forward, our focus will shift to upcoming projects from Reality AI, which promise to be even more advanced.

We sincerely appreciate all your positive feedback on Mark and how it has benefited you. Thank you for being part of this journey!

Introducing the Quanta Language Model Series by Reality AI - Make it Real.