Our AI Models
Discover our comprehensive suite of AI models, each designed to excel in specific domains and deliver exceptional performance for your applications.
Horus Family Models
A suite of high-performance open-source language models built for developers, researchers, and innovators across the Middle East and beyond.
Replica TTS Models
Multilingual voice synthesis supporting 20+ languages and 50+ dialects with natural-sounding output, designed for real-world applications.
STAM Optimizer
Stable Training with Adaptive Momentum — a JAX-based optimizer that dynamically adapts β₁ based on online gradient variance, achieving top-3 performance on 83% of benchmark phases.
ListenX STT Models
Advanced speech recognition models for accurate transcription across multiple languages and dialects, including specialized Egyptian Arabic support.
Traum Tokenizer
High-performance specialized tokenizer optimized for Flash-SLM with superior compression efficiency.
© 2026 TokenAI. All rights reserved.