Notes from the work

One email when a tutorial or migration write-up goes live. Nothing else, and an unsubscribe link in every email.

LeanZero

Two people in Romania doing Atlassian migrations, Forge apps and practical AI work for teams that would rather talk to the person doing the job. Most of what we learn ends up on this site.

Services

  • Atlassian migrations
  • Atlassian FastShift
  • Atlassian maintenance
  • Forge app development
  • AI development consultation

Topics

  • Jira
  • Jira Service Management
  • Confluence
  • Bitbucket
  • Atlassian Forge
  • Cloud Migration
  • Local AI
  • AI Coding
  • Certifications
  • Atlassian Team
  • All topics

Company

  • Blog
  • Tutorials
  • Contact

Community

  • Join Discord
  • Support this site

© 2026 LeanZero. All rights reserved.

Privacy policyTerms of service24/7 SupportTrust centerSecurityDPALegal notice
  1. Home
  2. Portfolio
  3. Elzi
  4. Features
  5. Nodes
Elzi feature guide

Nodes and strategies

The models Elzi hands work to, and the rules for which one does what.

A node is a model Elzi can hand work to, on your Macs or in the cloud. You make one by choosing what serves it: a provider, then a model. Chats run on a node or a strategy, and swarm builds run on a strategy or on your swarm pool. A strategy says which node does what: its roles, the order to try nodes in, and when to move to the next one.

Adding a node

  1. Set up a provider first: download a model in the LeanZero MLX tab, or connect a cloud provider with its key.
  2. Open Nodes in the sidebar and choose "New node".
  3. Pick the provider. For LeanZero MLX, choose the Mac, give the node a label and pick a model from the engine's models folder. For a cloud provider, pick the model.
  4. Set its weight. A higher weight gets a bigger share of the tasks in a build.

If no model is downloaded yet, the app says so and points you to the LeanZero MLX tab. You can also add a node from a chat with "Add a node to this chat…".

Nodes on other Macs

Today only this Mac's MLX node serves work. A node you save for another Mac waits in the pool until routing to other Macs ships. To use another Mac's model now, run it there through LeanZero Link.

Where nodes are used

  • Chats: "Run this chat on" picks a node, a strategy or "Any node (Auto)", which answers from your swarm pool. Delegates of a chat share its nodes. You can save a chat's nodes as a strategy.
  • Software builds: the build runs on your swarm pool or on a strategy you marked for builds.
  • Benchmarks: a Gauntlet run can enter your swarm instead of a single model.
  • Agent Work: the desk view shows which nodes are busy.

The Nodes row in the sidebar shows a node that is still loading or has failed before you send it work. When no node can take a message, the chat says so, and you can let it fall back to the next node.

Rules worth knowing

  • Each Mac has one LeanZero MLX node, because one engine serves one model at a time on a Mac. Adding the node loads and stops nothing; it sets the model builds use there, and chat follows the model that Mac is serving now.
  • Cloud nodes have no such limit.
  • LM Studio models running on your Macs join the pool by themselves.
  • OpenAI-compatible endpoints you add yourself work for chat, not for swarm builds yet.
  • Your other Macs appear once they join LeanZero Link.

Strategies

The Strategies tab next to Nodes holds your strategies. A strategy gives nodes roles (Planning, Build, Backend, Frontend, Testing and Chat), sets the order to try them in and says when to use the next one. Mark a strategy "Use for swarm builds" to make builds use it.

Related guides

ProvidersWhere models come from: the MLX engine on your Macs, or a cloud provider with your key.Software build sessionsOne brief becomes a planned, multi-agent build across your nodes.Agent sessionsChat with an agent inside a project folder, or set one up to work on a schedule.LeanZero LinkUse your other Macs: run a model there, copy models over, or split one across two.

Facts on this page were checked against our own test records on 6 November 2026.

Using Elzi

  • What Elzi is
  • Agent sessions
  • Software build sessions
  • Benchmark runs
  • Nodes and strategies
  • Providers

The MLX engine

  • The LeanZero MLX engine
  • Serving a model on one Mac
  • Distributed serving across two Macs
  • N-gram tables and prompt lookup
  • Speed: measured numbers
  • Memory lock

Media and devices

  • Images
  • Voices
  • LeanZero Link

Elzi

  • All feature guides
  • Download and user manual