How is the price of storage set in tokens per chunk?

I’d like to know how does the Autonomi network nodes come to resolution about how many network tokens to charge for storing one chunk of data?


I’ve already asked for that to be in the OP for each test/betanet. As far as i can tell they are going to do that @joshuef ?? It is after all one of the driving forces for people to run nodes. Not all but affects almost all, even if they would run a node for free.

While I have your attention @joshuef Another thing in the OP would be great, and is needed because of the changes.

  • store cost algo (as mentioned above)
  • max chunk size
  • max number of chunks

There is a reply to my asking in the current testnet topic with the formula and its simple enough but i’d have to graph it to get a visualisation of the effects. It seems that there is a crossing point at 60% full.

I’ll see if I can find it and post/edit it here


Here is a link to the current testing of store cost algo


The current Network price of storage, seems to be working like the ebb and flow of a water body. What I’m missing as a clueless observer in this fluctuation is the frozen water pockets. These would be what somebody is really willing to pay contrasted with what is offered for free in the storage market. Time will probably past before some start using the Network’s storage, especially if it makes no economic sense to them…

1 Like

A DEX function within the Autonomi Network is really the missing link imo.

The idea is the user running the safe client should, when operating within its own close space, (containing many close groups) see ‘close’ node GOSSIP updated aggregate price ask information aggregated in ‘close space’ DEXes advertising via GOSSIP latest journaled price info on what the ‘close space’ nodes are offering as cost of upload for each node in the ‘close’ space,

and within the safe client (user set once and forget) pre-set budget parameters will govern the range of automatically selecting which node offers they want to use…, meaning the proposed DEX service function needs to be an optional run option on safe node.

Like you I have returned recently to this project after an extended hiatus, which has me first on the forum in 2017, so I am still trying to re-assemble the complete ‘autonomi’ picture on what is really being rolled out.

Heck I just ran my first test node on a crap Win11 machine over the weekend to better understand what was going on, was partially able to get the node up, but never really got it across the goal line to earn rewards… (Day job gets in the way)



Currently we have “flow” but no “ebb” because chunk pricing is based on chunks on a node, but not “chunks the node is responsible for”. And nodes don’t delete chunks until they start getting full.

Next net will have a “ebb” mechanism for chunks to be released from the pricing mechanism and allow a node to reduce its prices as its responsibility is reduced (more nodes join the network in its group).


This is kind of what’s happening, in a way, but not exactly.

Group A has 5 nodes. Each member does their own price Cals based on their data fullness.

A5 just got put into the group because of churn and is more full than A1-A4. When a client asks for a store quote, it asks group A “give me quotes for this chunk”. A1-A5 respond, and A3 is picked and paid because it had the best quote.

Now, that may be slightly off because I think the client asks for quotes from the nodes with addresses closest to the chunk hash. And those may be in different groups. However, the premise is the same.

To the point of the price limit, Ive also advocated for that. You get 1 chunk that happens to hash to a “popular” address space and price skyrocket, I don’t want it to drain my account without asking. I and others have had some mighty expensive chunks stored.


I also thought in the past that a dex was super important, now i think not, it’s just a tool for kids to play.

Now I think AI is our most important tool, to solve real problems. The crypto space only idea now seems like create another coin and everybody gets rich, to me this is bull :poop:. If we are creating money tokens on the Autonomi Network we are just wasting time. It’s more important to cure diseases, set the world free with computational money and have fun.

This is not meant to be a disrespect or offense to @happybeing or others SAFERS, but some see £0.05

Others see 132,315,412 operations that can be executed by an AI, with the SAFE faucet every human on Earth could get a few nanos/day.

Michael Levin said it the best I think, I’m probably quoting this wrong :sweat_smile: when you’re measuring the intelligence of a system, it’s kinda like measuring your own intelligence.

About economic, when Storj offered 150 gb for free, for me that was a hint that my 1 nano is worth 150 gb. For free is now @ 1TB, my elementary skool teacher always taught me pictured beneath


“Eddy you can be a clueless consumer, butt please, thus please NEVER be a clueless spender”

But I think how much your willing to pay for storage on the Network ties into how much your willing to pay for a vps etc. This Network is our FIQ test…

SAFERS should be aware of paying 10K for a pizza…

In all honesty, even if some choose to devalue their token it doesn’t really matter the end game is paying ASI 1 nano for a operation

1 Like

Keep trying and dont be scared to shout for help.

I setup on the Alpha 5 network, its working on Win11, just one node, facuet worked, upload worked, but dead slow 3min 36sec for 528KB file that end up using 3 chunks when the notes stated chunk sizes were 512KB ? Seems it should have only used 2 chunks and then replicate later? Am I missing something?

Chunk size is a maximum of 512KB, but, if possible, it will always split into at least 3 chunks.


We need a minimum of 3 chunks to self-encrypt.
Try a larger upload and see how long that takes.

1 Like