Lite node

A Lotus lite-node is a stripped down version of a Lotus full-node capable of running on lower-end hardware. Lotus lite-nodes do not contain any chain-data and can only perform message signing and deal transactions. However, they are very quick to spin up and can process transactions in parallel with other Lotus lite-nodes.

Before we get started, let’s just go over the terms we’ll use in this guide:

LotusThe main Filecoin implementation, written in Golang.
Full-nodeA Lotus node that contains all the blockchain data of the Filecoin network.
Lite-nodeA Lotus node that does not contain any of the blockchain data of the Filecoin network. Lite-nodes rely on access to a full-node in order to run.


To spin up a Lotus lite-node, you will need:

  1. A Lotus node - For best results, make sure that this node is fully synced.
  2. A computer with at least 2GB RAM and a dual-core CPU to act as the Lotus lite-node. This can be your local machine. This computer must have Rust and Go 1.21.7 or higher installed.
  3. You must have all the software dependencies required to build Lotus.

Full-node preparation

If you have access to the full-node you’re using, you need to make some minor modifications to its configuration.

  1. On your full-node open ~/.lotus/config and:

    a. Uncomment line 3. a. Change to

    ListenAddress = "/ip4/"

    Save and exit the file.

  2. Create an API token for your lite-node to use:

    # lotus auth create-token --perm <read,write,sign,admin>
    lotus auth create-token --perm write

    Which permissions you choose will depend on your use case. Take a look at the API tokens section to find out more →

  3. Send this API token to your lite-node or to whoever will be the administrator for the lite-node.

  4. If you have the lotus daemon running, stop it and start it again. This forces Lotus to open the API port we just set.

Next up, you’ll create the Lotus executable on your lite-node and running it in lite mode!

Create the executable

You need to create the Lotus executable to run your lite-node with. This process is the same as when creating a full-node.

AMD and Intel-based computers

  1. On the computer that you want to run the lite-node from, clone the Lotus GitHub repository

    git clone
    cd lotus
  2. Create the executable, but do not run anything yet:

    make clean all
    sudo make install

    If you run into errors here, it may be because you don’t have all the Lotus dependencies installed. Take a quick look at the Lotus installation guide and double-check that you have all the dependencies installed, along with Golang and Rust.

  3. Move onto starting the lite-node.

M1-based Macs

Because of the novel architecture of the M1-based Mac computers, some specific environment variables must be set before creating the lotus executable.

  1. Clone the Lotus repository from GitHub:

    git clone
    cd lotus
  2. Pull-in the submodules:

    git submodule update --init --recursive
  3. Create necessary environment variable to allow Lotus to run on ARM architecture:

    export GOARCH=arm64
    export CGO_ENABLED=1
    export LIBRARY_PATH=/opt/homebrew/lib
  4. Move into the extern/filecoin-ffi directory and checkout to the m1-portable branch:

    cd extern/filecoin-ffi
    git fetch -a
    git checkout releases
  5. Create the filecoin-ffi executables:

    make clean
  6. Move back to the root Lotus directory and create the lotus daemon:

    cd ../../
    make lotus
  7. Move onto starting the lite-node.

Start the lite-node

You’ve got the Lotus executables ready to go, and you have access to a Lotus full-node. All that’s left is connecting your Lotus lite-node to the full-node!

  1. On the lite-node, create an environment variable called FULLNODE_API_INFO and give it the following value while calling lotus daemon --lite. Make sure to replace API_TOKEN with the token you got from the full-node and YOUR_FULL_NODE_IP_ADDRESS with the IP address of your full-node:

    FULLNODE_API_INFO=API_TOKEN:/ip4/YOUR_FULL_NODE_IP_ADDRESS/tcp/1234 lotus daemon --lite
    2021-03-02T23:59:50.609Z        INFO    main    lotus/daemon.go:201     lotus repo: /root/.lotus

    The API_TOKEN variable must be followed by a colon sign (:).

    If you don’t have an API_TOKEN, you can run the above command without one and just gain read-only access to the full-node:

    FULLNODE_API_INFO=/ip4/YOUR_FULL_NODE_IP_ADDRESS/tcp/1234 lotus daemon --lite

    FULLNODE_API_INFO variable accepts both Multiaddr and Canonical (RFC1918, RFC1034) network address formats. For example, for Glif nodes the Lotus Lite run command may look like:

    FULLNODE_API_INFO=wss:// lotus daemon --lite
  2. You can now interact with your Lotus lite-node:

    lotus wallet balance f10...
    100 FIL

A lite-node is limited in what it can do and is designed to only perform message signing and transactional operations. Lite-nodes cannot seal data or query the chain directly. All chain requests go through the attached full-node. If for whatever reason, the full-node goes offline, any lite-nodes connected to it will also go offline.

Access and permissions

Setting up a Lotus lite-node without using an API token from a full-node results in the lite-node having read-only access to the full-node. While read-only access should be fine for most use-cases, there are situations where you need write access to the full-node.

Use cases

A Lotus lite-node can perform transaction-based functions like creating the transaction, proposing deals, signing messages, etc. They do not have any chain data themselves and rely on a full-node for chain data completely. Lotus lite-nodes are completely useless on their own.

One use case is a service that needs to sign multiple messages a minute, such as an exchange. In this case, the service could have multiple lite-nodes specifically to sign and deal with transactional computation, while a single full-node maintains the chain data.

Another scenario is an organization with a lot of data they want to store on Filecoin but no resources to run a full-node. They could create a Lotus lite-node to create deals with a full-node that they trust, and then once a miner has been found, the lite-node can transfer the data to the miner.

Benefits and drawbacks

Since Lotus lite-nodes do not need to sync any chain data, they’re able to spin up quickly. Due to their minimal hardware requirements, it’s possible to spin up multiple lite-nodes with quite a small footprint. A project can horizontally expand or shrink its processing power by adding or removing Lotus lite-nodes whenever necessary.

However, lite-nodes come with some significant drawbacks. While there are some functions that a lite-node can perform without access to a full-node, all transactions and processes that require access to chain-data must be routed through a Lotus full-node.

Edit this page on GitHub