Skip to content
Back to Articles
August 2026Cortiqa Research

Engineering Sovereign AI: The Menothus Architecture & Falin-300M

How we engineered a 297M parameter sovereign Small Language Model (SLM) from scratch in India, featuring 8:1 GQA and Parallel SwiGLU Attention.

#Research#Models#Architecture

Building foundational artificial intelligence architectures from scratch requires reimagining every layer of the compute stack.

With Falin-300M, Cortiqa set out to demonstrate a sovereign proof-of-concept: that efficient, high-performance language models can be conceived, structured, and trained entirely natively in India.

Why Small Language Models (SLMs)?

While frontier models exceed hundreds of billions of parameters, real-world deployment on edge devices, local servers, and browser extensions demands compact footprints with near-zero latency.

A 297-Million parameter model can be deployed on standard consumer GPUs, mobile CPUs, and embedded hardware without demanding massive cloud infrastructure or high thermal budgets.

Key Architectural Innovations

Falin-300M is built upon Cortiqa's proprietary Menothus architecture:

  1. Extreme Grouped Query Attention (GQA): Utilizing a 16:2 head ratio (8:1), saving 75% of KV-cache VRAM during inference.
  2. Parallel Attention + SwiGLU FFN: Executing attention mechanisms concurrently with feed-forward computations to maximize GPU throughput.
  3. Hybrid Sliding Window Attention (SWA): Balancing local context (512 tokens) with dense global causal attention every 4th layer.
  4. Native FlashAttention / SDPA: Clean PyTorch scaled dot-product attention without fragile third-party compilation layers.

Open Research Preview

Falin-300M is available on Hugging Face for academic evaluation, non-commercial research, and domain fine-tuning.