❓Fast Worker FAQ

Have a question that isn't on this list? We're always updating the FAQ, just give us a ping on one of the channels listed at the bottom of the page 🤙

How do I access my Fast Worker earnings?

Unlike the Relaxed Worker network which automatically sends token earnings to the Sogni Account directly, with the Fast Worker you must Claim Worker Earnings periodically. Proceed to https://nft.sogni.ai/myearnings to do so.

My GPU isn't on the list, when will you support it?

The Sogni Fast Supernet strives to provide super fast and reliable render times (often less than <5 seconds) to artist no matter if returning a single image job or a batch of 16 images. This unique parallel processing capability is only as fast as the slowest worker in the batch. So in the current testnet phase we are restricting the Fast Supernet to these, top of the line but very popular, residential GPUs.

In the future we will be releasing new worker job types and Supernet speed tiers which will allow us to accommodate more GPUs. Our end goal is to allow all GPUs to participate.

How do I get the Fast Worker NFT and API Key?

Just download the Fast Worker installer for your OS and follow the steps in the interactive installer. As part of that process your NFT will mint directly to your Base Sepola Testnet wallet, the wallet that was set up automatically for your Sogni account. During this process you'll also be able to create an API Key for your account. Both are needed to complete the setup for your worker.

The NFT_TOKEN_ID must be supplied because you may have multiple NFTs under your address and the Supernet needs to know which one to pull and build reputation against for your worker.

Can I run more than one Fast Worker?

Yes you can run as many as you have the GPUs to allocate for so long as there are still unclaimed Fast Worker NFTs available to mint are made available in batches that help ensure the supply of workers can go up in-line with platform demand. This helps ensure maximum job rates (and token rewards) for the number of GPUs on the network.

Can I run multiple Fast Worker instances from one machine that has multiple GPUs?

Yep this is supported though each Fast Worker must be configured with it's own NFT TOKEN ID even when running from one machine. You must also have a minimum of 30gb of ram for EACH instance of Fast Worker you are running. See details in Advanced Configuration.

The site nft.sogni.ai is not letting me mint more than 1 Fast Worker NFT ?

Site restrictions currently only allow Sogni accounts to mint an NFT if the address is not already holding one to prevent hoarding while minting is free during Testnet. However you can set up a new Sogni account for each worker, even using the same email address .

How does the Leaderboard and airdrop rewards work?

The Fast Supernet Leaderboard is a live rank of Fast Workers by how many tokens were earned against the NFT it was logged in with. Tokens are earned by completing AI inference jobs on Sogni's Fast Supernet. Job analytics and reputation are stored against the NFT and not the Sogni account or address for this leaderboard ranking. This means when you run multiple Fast Workers, even from the same Sogni account, each will have a separate entry on the leaderboard.

An airdrop of Sogni's Mainnet token will be distributed to the Base network wallet automatically set up for each Sogni account with exponentially higher rewards based on leaderboard position. The final testnet leaderboard snapshot is tentatively end-of-March 2025 for an early April airdrop.

Can I run both Relaxed / Mac Worker(s) and Fast Worker(s)?

This is fully supported but it is recommended not to do so from the same Sogni account.

Since the Relaxed Worker leaderboard is a single leaderboard for the entire Sogni address / account (with support for multiple Mac Workers) and Fast Worker jobs also contribute to address analytics, the leaderboard mechanics automatically disqualify and remove an account from the Relaxed Leaderboard if a Fast Worker authenticates with that account's API Key.

If your Relaxed Worker is automatically removed in this way you can reach out on Discord or [e-mail] and we may be able to restore it for you.

Can I switch my worker from Stable Diffusion to Flux and vice versa without reinstalling everything?

Yep! Just re-run the installer script. It will detect that you have already completed setup and only ask you to pick which worker image you want to switch to. All models are cached locally in the /data-models/ folder so you won't have to redownload anything. You can delete that folder at any time to save space. Worker analytics stays with the NFT so you could even switch machines and keep your spot on the leaderboard.

How can I update my workers to the latest version of the Fast Worker software?

Sogni Fast Worker is designed to automatically check and pull the latest worker image from Docker Hub every time it starts and at a periodic interval when it is running so you normally don't need to do anything. To expedite an update you can simply stop and start your worker again to trigger the update check.

My Fast Worker won't auto-update and got disconnected for running an unsupported version, how do I fix it?

If you're running Fast Worker on a 3rd party DePIN provider like Salad.com you may need to occasionally manually "refresh" the docker image being pulled. You can do this by simply clicking "Edit" on the Container Image Source URL and re-entering the same path that was previously there, for the service to pull and cache the latest version of the "latest" tagged docker image. For Sogni Fast Worker Stable Diffusion: sogni/sogni-stable-diffusion-worker:latest For Sogni Fast Worker Flux: sogni/sogni-flux-worker:latest You may also replace the ":latest" tag above with the actual latest numeric version of the worker image as listed on our public Docker Hub pages for our Stable Diffusion worker and Flux worker. You may also hit these corresponding utility endpoints to return the latest Stable Diffusion and Flux verion paths.

Can I use my computer while running Fast Worker?

Casual use of your computer while running Fast Worker is completely fine, however if you use any applications or games that may leverage your graphics card it may cause your worker's performance to drop. If your machine is too bogged down to perform jobs quickly it will be significantly deprioritzed in the system to where it will get less jobs. So if you plan on using your computer for anythign intensive, it's best to shut down Fast Worker and turn it back on after you are done.