What this means for Node operators
The new pricing model will not change your per-GB storage and egress payout rates. New offerings are being created that are intended to match customer use cases. Global Collaboration is for performance sensitive use cases and situations where customers share data globally. The Regional Workflows tier is designed to serve customers with geographic or compliance constraints on their data. The Active Archive tier will serve those looking for an affordable “hot” archive solution.
One of the ways Storj plans to improve the performance of the Global Collaboration tier is by increasing the expansion factor of the data stored there. As the utilization of that product increases, the space consumed on nodes should accelerate faster than in the past. Also, Storj plans to use nodes provided by the community to support workloads on all future pricing tiers, including the regional tier. If the customer’s needs (i.e. data needs to be stored in EU) can be met by community nodes, the plan is to use those for the capacity needed.
If you are also a Storj user, you should have received an email from Storj explaining that existing projects will remain at legacy pricing for a full year.
Thank you for your continued support in powering Storj’s distributed network!
So… it looks like the new offering is more expensive, which will be good from the storj profitability perspective.
Question, there is the mention of latency. What is storj’s time to first byte like, and are there worthwhile changes to try to make it faster?
Is there anything that can or should be done from the node operator side to help latency? For instance, data could be stored on SSD. Actually, that’s really the only change a node operator could make because everything else is pretty much plug and play. But the satellites or overall platform could try to select faster nodes more aggressively or something…
It seems like this could spike another “potato nodes vs bigger rigs” competition. The faster your node and connection (low latency) is, the more races you win
On Storj Pricing | Seamlessly scale on the cloud built for media production, the “Global collaboration” tier currently shows Egress: Included. However, when you toggle the “Is egress included in all tiers?”, it explains that only the amount of egress equal to the data stored is free per month. To avoid confusion, it would help to either add an asterisk or replace “Included” with a very short clarification.
Just as an idea: Since the satellite knows what tier of storage is being used, I would also be fine with tiering how much my storagenode gets paid. For example, in the “Global collaboration” tier, if I received a little more per TB, I would be happy to get paid much less, or even nothing, for egress. This way you could even offer 10x, 100x, or potentially unlimited egress.
One small concern: There has been some discussion that if egress were completely free, some storagenodes might throttle it, which would be problematic. However, since there is also paid “Active archive” egress, storagenodes may not know whether they are being paid for egress or not. This already discourages throttling, since I would not want to miss out on potential earnings.
There are gives and takes on this. The segment fee drove revenue and while eliminating the segment fee removes a barrier to adoption as it was unique to Storj.
200-600ms is our time to first byte depending on where you are in the world. us-select-1 is able to do much better than that for workloads that are local. As part of increasing expansion factor, we are working on bringing that time to first byte down around the globe.
With the new hashstore rollout, we don’t believe using SSD will have significant improvements. Probably the biggest thing is monitoring your ISP’s upstream bandwidth and making sure there is headroom so it doesn’t get saturated.
As discussed in the forums previously, there is already increased preference for performant nodes than in the past. In terms of ROI on rigs, there is definitely going to be a dropoff on what makes it worthwhile to squeeze a few ms more out on the storage node side. Most likely node location and ISP connectivity is going to dominate after that.
This looks like an exciting step. I like the concept of offering different tiers of storage. But it’s a little (perhaps intentionally) unclear how that translates into technical implementation. I’d love to hear a bit more about how the different tiers translate to RS numbers and segment size, etc. Or any other technical differences in perhaps node selection or other aspects. What differences make the global collaboration tier better for performance and the active backup tier better for long term hot backup storage?
I’m also curious what existing plans will eventually be migrated to?
As well as how this change impacts global network vs select?
This is definitely something we’ve discussed as a team in the past as there is competitive pressure on including more free egress and our current model of paying SNO’s a fixed $/TB limits our options. None of those discussions led to a proposal that we were happy enough to send to the community to get further feedback and even consider changing… you identifed one concern/issue and there are others.
If I could wave a magic wand and have anything possible then compensating on some measure of throughput capacity is what I believe best captures the network needs and aligning to SNO cost structure.
Great questions, I hope to be able to share more in the coming weeks.
This was covered in the linked FAQs in the emails that were sent to customers. If they don’t choose a different tier by November 1st, 2026 they will be migrated to Active archive.
Select continues to be the offering for compliance/data sovereignty and in rare cases when TTFB performance in the U.S. East coast drive the buying decisions. Our global SNO network continues to be a huge competitive advantage and the new Global collaboration leans in on that advantage. Customers love the consistent performance anywhere that you can only get with such a diverse supply of nodes, networks, etc. and we are doubling down on this.
As a customer who has been using Storj for a few years to back up my home and personal business data, with total expenses of about $15-20 per month, I will have to adapt, yes
With the new prices, I would rather switch to my own rented S3 server (even 2 for safety) in a year than continue using Storj. With this new plan, my costs will increase 2-3 times, maybe even more, due to storage and minimum object size. The proposed plans do not cover my usage model at all, and I am not ready to pay THAT MUCH for object storage. And I am not even talking about the constant problems with data latency. And about the (not constant) availability of Storj IP addresses in countries where I am constantly on business (China, or recently, suddenly Spain). Somehow this service does not pull the title of the fastest, most accessible and distributed storage for me.
How will this affect the operators’ income? I hope, guys, that you will really earn more in the long term than you do now. But I wouldn’t count on it too much. You’ve almost lost me - a small client.
The minimum object size cost should be less than the existing segment fee in every case.
Can you share more about your usage details or send the details or account email to me directly? I’d love to take a look and understand your use case more.
Both mostly largely out of our control unfortunately.
For now it is mostly cold archive storage for my data with mostly ingress traffic. Some deletes of obsolete/stale data, probably about 100-200GB of deletes per month. (I’m still currently estimating how much and how often I delete). Storage size is 3.3 TB, ~200K segments, egress is varying from 1-4 GB normally, up to 0.5-1TB in case of backup restores / downloads.
My current invoices are following: ~13.5$ for storage, ~1.75$ for segments, the rest is for egress (up to 15$ extra, but typically around couple of $ per month).
According to my quick estimate, with my current usage scenario, my invoice will increase by about two times in a year when the new tariffs come into effect. That is if you don’t introduce any other new fees. I haven’t calculated exact expenses increase yet, but my storage costs are still increasing significantly. Much more inflation in my country. Thank you for giving me a year to get away from you. Although what to do with the money from the Storj tokens that remain on my account balance is a good question.
I planned to scale bigger in a couple of years, up to 8TB of storage, but now it looks more cheaper to buy 2 different cheap storage VPS and host some replicated network filesystem on top of it, and I not mentioning increase of access-times.
I could imagine that if storj increased the expansion factor there would be more nodes competing for any individual upload and definitely download, which would lead to more canceled transfers after the transfer is completed from the faster nodes.
So in otherwords if the VPN slows the node, it will be punished more.
