mesh-llm

Health

Risk
risk 0 ok
Version
0.80.0 no newer release in its git tagsrepology ↗nixkeeper ↗
Update bot
no attempt logged · not in the queue
Channels
master 0.80.0 · nixos-unstable absent · nixos-26.05 absent no channel is behind master
Advisories
none open
Packaging
meta complete
Description
Pool spare GPU capacity to run LLMs at larger scale
Long description
MeshLLM pools the GPUs of several machines to serve large language models through an OpenAI-compatible API and a web console. By default it bundles the Metal runtime on macOS and the CPU runtime on Linux; GPU backends on Linux are enabled with the nixpkgs settings `cudaSupport` (unfree) or `rocmSupport`, or Vulkan by overriding skippy-native-runtime with `vulkanSupport = true`.
License
Platforms
3 platforms: Linux, Darwin · not on pattern
Main program
mesh-llm
Source provenance
not set: built from source
Outputs
out
Status
not broken, not insecure, free
Maintainers

Version bumps 0

all open bumps
None open.

Pull requests 4

mesh-llmPlugins.{metrics,flash-moe,blackboard,agents}: init draft
#583566 by kleinbem · opened 2w ago · checks pass
skippy-tools: init at 0.80.0
#583562 by kleinbem · opened 2w ago · checks pass
nixos/mesh-llm: init module
#581183 by kleinbem · opened 2w ago · checks pass
Closed or merged in the last 60 days 1

Issues 0

None open.