Kalsa
Engines / Lidenburg / Lidenburg/llama.cpp

Lidenburg/llama.cpp

NVIDIAForkMIT

Adds an expert cache with GPU and RAM tiers to llama.cpp on NVIDIA GPUs. It is a fork of ggml-org/llama.cpp.

Benchmarks

  • On RTX 5080 (16GB) + 32GB RAM with Qwen3.6-35B-A3B (UD-Q4_K_M), it decodes at 80 tok/s, against 50 tok/s for expert cache off (+60%). Source Note: Approximate tok/s as published; 256k context

Listing

Listed: Passed all seven checks.

Security review

Not reviewed yet.

Install

To install it, follow the project's own instructions on GitHub. The engine runs with your user's permissions. macOS doesn't sandbox it, so install only engines you trust.

Something wrong or out of date? Write to help@kalsa.io, or ask us to remove this page.