37 tok/s from Qwen3-Next-80B (a 48 GB model) on a laptop with 32 GB of memory, 24 without drafts, where llama.cpp manages 2.8; reading a prompt, 62 tok/s against 16. Janas-LLM was faster than llama.cpp on every model compared on the same laptop (Debian 13 Linux, Intel Core Ultra 9 185H, integrated Arc GPU, 32 GB, NVMe SSD; 9–11 October 2026).
Janas is a family of tools for artificial intelligence, written in C with no dependencies beyond the C library. Its first component, Janas-LLM, runs large mixture-of-experts language models on ordinary computers, with or without a GPU: the weights stay on disk, the experts each token needs are streamed from an NVMe SSD, and the ones that keep coming back stay in memory. An 80-billion parameter model runs, and answers, on a laptop with 32 GB of RAM.
Designed and written by Maurizio “camauri” Cammalleri.
Status
This is an experiment, not a product. Early development; public API changes without notice. The engine deliberately stresses hardware (fills free memory with expert cache, O_DIRECT disk reads). Developed primarily on one machine; janas-bench is recommended for other hardware reports.
Benchmark highlights (Q4_K_M vs llama.cpp)
| Model | reading 1,024 tokens: Janas | llama.cpp | writing: Janas | llama.cpp |
| Qwen3-Next-80B-A3B (48.4 GB) | 62 tok/s | 16 tok/s | 24 tok/s | 2.8 tok/s |
| Qwen3-Coder-Next (48.4 GB) | 66 tok/s | 14 tok/s | 22 tok/s | 2.2 tok/s |
| Qwen3.5-9B | 111 tok/s | 78 tok/s | 14 tok/s | 8.7 tok/s |
Models are distributed pre-converted on Hugging Face under prabanta-dev. Linux x86-64 and Windows 10/11 builds available via GitHub releases.