2026•Technology Innovation Institute (TII)
Falcon H1R-7B v1.0
Falcon H1R-7B is a 7-billion parameter reasoning-focused language model developed by TII in Abu Dhabi. It combines Transformer layers with Mamba2 state space components, enabling efficient processing of sequences up to 256,000 tokens. The model is fine-tuned for tasks in mathematics, coding, and general reasoning, achieving benchmark scores that rival larger models.
Hybrid Transformer and Mamba2 architecture
Supports context lengths up to 256,000 tokens
Fine-tuned for math, coding, and general reasoning tasks
Version1.0
RegionUnited Arab Emirates