A local, single-process, multi-GPU-capable AI inference runtime written in C++23/CUDA — text, vision, audio, image generation, embeddings, RAG, and multi-agent workflows from one executable.
Proprietary — All Rights Reserved. Copyright © 2026 Boominathan Alagirisamy. This software is not open source; no use, copying, modification, or distribution is permitted without prior written permission. See LICENSE.
125 commits
C++
71.4%
TypeScript
11.4%
Cuda
7.5%
Python
6.4%
Shell
1.2%
A local, single-process, multi-GPU-capable AI inference runtime written in C++23/CUDA — text, vision, audio, image generation, embeddings, RAG, and multi-agent workflows from one executable.
Proprietary — All Rights Reserved. Copyright © 2026 Boominathan Alagirisamy. This software is not open source; no use, copying, modification, or distribution is permitted without prior written permission. See LICENSE.
125 commits
C++
71.4%
TypeScript
11.4%
Cuda
7.5%
Python
6.4%
Shell
1.2%