Upcoming onicai SNS Decentralization Swap

Developers Showcase
,
1

image
image937×194 14.6 KB

NOTE: an updated SNS proposal was announced on December 22, 2025 in this reply: Upcoming onicai SNS Decentralization Swap - #24 by icpp

Today we’re excited to announce our plans to propose to you the onicai SNS! After laying many technical foundations for decentralized AI on the IC during the last couple of years, we are convinced that the timing is right to hand over the control to the community to achieve decentralization of AI on the governance level as well. We’re grateful to the community which has supported and formed around the different onicai products and accomplishments, and cannot wait to see where we’ll take this together next as the onicai SNS!

This post contains a summary of the onicai SNS Whitepaper, a link to the full whitepaper, as well as a link to the parameters we’re proposing for the decentralization sale and onicai SNS.

onicai SNS Whitepaper: https://www.onicai.com/files/onicai_SNS_Whitepaper.pdf

GitHub repo with the file onicai_sns_init.yaml containing all proposed parameters:

We’re looking forward to your questions and feedback!

Arjaan (@icpp), Nuno (@Nuno ), Abhishek (@kanodia ), Patrick (@patnorris )

Executive Summary

ON Internet Computer Artificial Intelligence (onicai) is building a decentralized Intelligence-as-a-Service layer on the Internet Computer, with the onicai SNS as its community-owned governance framework. The long-term vision is a hybrid network of sovereign AI agents, tools, and services that operate autonomously and trustlessly on behalf of individuals and organizations—distributing intelligence to the many by the many, rather than concentrating it in a few centralized providers.

Technically, this is driven by two core concepts. First, Proof-of-AI-Work is a protocol that coordinates decentralized AI agents, verifies their work cryptographically, and rewards valuable contributions in a trustless way. Trustless means the system operates through decentralized code and cryptography, without requiring users to trust any central authority. Second, Chain Fusion AI allows “Hybrid Agents” to span on-chain coordination and governance, on-device private workloads, and off-chain heavy computation in Trusted Execution Environments (TEEs) or via verified APIs. Together, they enable modular, interoperable AI ecosystems where agents can be spun up on demand to handle specific tasks without the “context bloat” of monolithic AI workflows.

The onicai SNS will immediately own and govern a substantial portfolio of existing, production-tested assets: the Proof-of-AI-Work protocol and its first live app funnAI, the IConfucius on-chain agent, the DeVinci decentralized chat app, ICGPT and the llama_cpp_canister for on-chain LLMs, the Charles AI NFT project, the icpp-pro C++ canister development kit, and several hackathon-winning applications including a Bitcoin Donation demonstrator. All relevant smart contracts and repositories will be consolidated under the SNS, open-sourced, and thus brought under community control.

Governance and development funding are structured around a milestone-based, proposal-driven treasury model. The SNS treasury will be funded by the decentralization sale, product revenues, client payments, liquidity fees, and staking yields and is fully under token-holder control. Development service providers (including the current team) request funds in two tranches per milestone (50% before and 50% after delivery), with work and outcomes vetted by the community. The majority of the ICP and ONICAI tokens in the treasury will be staked at genesis to unlock gradually and enforce capital discipline with long-term sustainability.

The ONICAI token is both a governance and utility token. Staked ONICAI grants voting power over the SNS roadmap and key decisions. Within the onicai DeAI Platform, ONICAI will power access to premium features, act as escrow and trust capital for listed solutions, serve as a settlement asset for payments between users, agents, and service providers, and influence solution placement and discovery. Rewards and bounties will be issued to align contributors with long-term network growth.

The onicai DeAI Platform will act as the main gateway to decentralized AI on the Internet Computer by integrating existing apps (funnAI, DeVinci decentralized chat app, etc.) and future solutions from both onicai and third parties. It will support sovereign Proof-of-AI-Work setups for individuals and organizations, multi-model DeAI chat (on-device, on-chain, off-chain), and a growing catalog of tools, agents, and AI experiences, compatible with emerging standards such as Model Context Protocol (MCP) and Agent2Agent where relevant. The SNS’ complementary B2B AI Development House will use the same stack to deliver custom DeAI projects, feeding revenue and additional capabilities back into the SNS.

Led by an experienced global team and supported by partners such as Outlier Ventures, DFINITY, and several ICP Hubs, onicai enters its decentralization phase with real users, live products, and a clear roadmap. It is positioned to become a foundational layer for sovereign, decentralized AI on the Internet Computer and across interconnected Web3 ecosystems.

Key Tokenomics Parameters

Parameter Value
Total Token Supply 1 Billion ONICAI
Minimum Participants 100
Minimum Direct Participation 80000 ICP
Maximum Direct Participation 300000 ICP
Minimum per Participant 1 ICP
Maximum per Participant 20000 ICP
Participant Vesting Schedule Interval 3 months
Participant Vesting Schedule Events 8

Token Allocation at SNS genesis

Category Percentage of total ONICAI Supply
Decentralization swap 20%
Developer Neurons 13.5%
Advisor: Outlier Ventures 2%
SNS treasury 64.5%

Decentralized Assets

The Development Team will transfer ownership of the following projects to the SNS:

Project Description
Proof-of-AI-Work The innovative protocol to control and coordinate AI agents by merging AI and decentralized tech. The onicai team developed the concept and implemented the core technical building blocks for running Proof-of-AI-Work applications on the Internet Computer.
funnAI Launched on the Internet Computer on June 28, 2025, this gamified mining protocol is the first-ever Proof-of-AI-Work application, developed and productized by the onicai team. Responsibilities cover development, maintenance, AI service provision, monitoring, support, community management, marketing and communication.
IConfucius The first ever fully on-chain AI agent running on the Internet Computer featuring the integrated ICONFUCIUS rune created on odin.fun
DeVinci The end-to-end decentralized AI chat app hosted on the Internet Computer and running a wide selection of state-of-the-art Large Language Models on the user’s device (laptop and smartphone) via common Web browser APIs.
ICGPT The first ever on-chain Generative AI that leverages a custom port of the llama.cpp inference engine to run open-source Large Language Models as canisters on the Internet Computer.
Charles The first ever on-chain AI NFT that lets owners dynamically update their NFTs with on-chain Generative AI running as canisters on the Internet Computer.
icpp-pro The Canister Development Kit to write smart contracts for the Internet Computer in C++
Hackathons related work The onicai team has developed several solutions for hackathons with a focus on decentralized AI. All related creations and repositories are handed over to the SNS.
2

For ppl who doesn’t know you could maybe link some videos like the one with Kyle(The Swop) it should spread awareness. Other than that you have our full support.
Also we can not thank you enough for your contribution to IC. Who knows knows and bows. :wink:

icpp-pro is one of the OG for IC. We used and referred to it many times even with the " Install didc" part included in proposal reviews.

Again HATS off Thank you and God bless !

3

What is the relationship between onicai token and funnai token? Will there be any airdrops based on the amount of funnai token held?

4

I saw BOB die with my own hands when someone made the decision to make Alice. So what did it do? It took an already stretched thin ecosystem populace and diluted it even more. Now “support BOB? Buy Alice? Buy BOB?” Literally cut Buy pressure in half. For what? Literally nothing. No benefit.

What do we lose by simply making an SNS with funnai tokens? Send funnai to the nns, stake it to the NNS and viola. Vote, govern, make yield from the tokenomics engine in funnai. We only make funnai stronger/more bullish and DONT dilute our purchasing power. If an “onicai” token exists then funnai gets half the buying pressure since the other half (from core supporters) goes towards onicai token buying. This leads to funnai being financially unattractive, AND AS A RESULT onicai being unattractive. Why do that? Why not have funnai itself be the governing token? This makes the most rational sense. Need a treasury for the team? Make a funnel from the maining operations, 1% towards the treasury or something. That can be enacted at the protocol level and solve all issues related to funding or management of funnai.

To create a separate onicai SNS is a pure loss/loss situation and IMO a backstabbing to the community who through funnai have been supporting onicai however they can.

Im strongly opposed to an onicai SNS, and encourage using funnai as the token for the SNS. This is both fair to investors now, and also more bullish for investors in the future.

5

Thanks for this announcement, it looks exciting. I have a range of questions.

Canisters

Could you list out the jobs of each of these canister please? Could you also point to a means for the community to verify that in each case?

  • dpljb-diaaa-aaaaa-qafsq-cai
  • dikpv-oqaaa-aaaaa-qafsa-cai
  • x6occ-biaaa-aaaai-acqzq-cai
  • xzpew-mqaaa-aaaai-acqza-cai
  • 4v3v2-lyaaa-aaaag-abzna-cai
  • otmmw-3yaaa-aaaag-ab2na-cai
  • 4c4bn-daaaa-aaaag-abvcq-cai
  • ounkc-waaaa-aaaag-ab2nq-cai
  • 6uwoh-vaaaa-aaaag-amema-cai
  • 6tht4-syaaa-aaaai-acriq-cai
  • 6ugvi-7aaaa-aaaai-acria-cai
  • 4j33a-miaaa-aaaai-acrhq-cai
  • 4o25u-bqaaa-aaaai-acrha-cai
  • 5rsod-ciaaa-aaaai-acrdq-cai
  • ekral-oiaaa-aaaag-acmda-cai
  • fj5jn-2qaaa-aaaag-acmfq-cai

Tokens

Could you also clarify the relationship between the FUNNAI token and the ONICAI governance token. What is it that gives the FUNNAI token value for holders? I’m asking largely because the value for ONICAI token holders is somewhat vague, and I’m wondering how it compares in terms of utility.

Funding

Could you elaborate on this? Presumably this refers primarily to ICP in the SNS treasury, raised by the SNS sale? Could you post the milestones?

Voting Power

This is what voting power at genesis will look like.

image
image456×377 9.67 KB

The developers have close to 50%, and that’s with their neurons on average having about a 12 month dissolve delay. The max is 24 months, and the dissolve delay bonus is 100%, so after the vesting period ends (12 months), if the dev team wanted to they could crank up their dissolve delays to 24 months on all their neurons and would have well over 50% of the total VP.

I appreciate that lots of brilliant work has gone into engineering these products, which are cutting edge. Nonetheless, I’m not sure what to think about this sort of VP distribution. Essentially it gives the developers the option to seize control of everything 12 months after launch. I don’t think SNSs should make eventualities like that easy or practical.

Thanks in advance

6

FYI - thank you all for your questions and feedback.
Want to just mention we’re preparing the answers and will post them shortly.

7

Thanks Onicai team,

I wonder

why SNS now? How the current sluggish crypto space could affect fundraising?

What role funnAi will play as a subsystem in the entire Onicai deAi platform in the years planned on the roadmap? In % terms, how would your rate its importance? What rewarding rate would be appropriate for funnAi mAiners operators?

Any financial plans and projections, based on milestones to be delivered or in quarterly basis?What are the funding requirements for each milestone?

What do you think on multiple rounds of fund-raising after each milestone delivery?

How do you justify 20% décentralisation sale for 300k ICP max, which gives an evaluation to Onicai project of $1.5M-5M mCap while funnAi itself has a current 125k mCap? (A financing leverage of 10x to 50x)

What are some of they key risks the team forsees when implementing Onicai growth plans?

Best, E

8

Why not include BOB? Besides, BOB ranks among the top on IC in terms of both trading volume and popularity.:thinking:

9

Hi,

Before I can post a review:

Full canister documentation — What does each canister do? What’s the controller? What’s the upgrade path?

Voting power analysis at various time horizons — Show the math for VP distribution at 6, 12, 18, 24 months

What is the relationship between FUNNAI and ONICAI? Can holders convert? What happens to FUNNAI after SNS launch?

Will team members, advisors (including Outlier Ventures), or affiliated parties participate in the swap? If yes, what is the maximum combined ICP?

10

Assuming you make it through the pre-SNS forum discussions alive - I wish you good luck in the SNS! I will be participating! :heart:

11

So youre saying i dont need to learn rust? Amazing. How did i not know about this?

20% thats like 10x the standard offer these days!

You have my vote good sir!

12

Hi everyone,

First, thank you for taking the time to read the whitepaper, share your feedback, and ask thoughtful, constructive questions.

We intend to address all of them so you can make a well-informed decision.

Before doing so, however, we will first finalize a key design decision: whether to proceed with the two-token model (ONICAI + FUNNAI) as outlined in the current whitepaper, or to adopt a single-token model using ONICAI only, with the funnAI application refactored to operate exclusively on the ONICAI token and the FUNNAI token deprecated.

This decision will be made by Friday, December 19. Once finalized, we will provide detailed responses to all questions and share an updated whitepaper along with a revised sns_init.yml for your review.

13

Overall, the SNS setup looks solid, and it’s clear that a lot of thought went into the neuron distribution, kudos for that.

That said, in addition to the already mentioned voting power concerns, I have a few additional observations:

  1. I noticed that the rejection cost is set to 100 tokens. With a fully subscribed sale, this implies a token price of roughly 0.0015 ICP per token, which at today’s prices translates to about $0.50 per rejected proposal. In my opinion, this is quite low and may not be sufficient to discourage spam or low-effort proposals.

  2. This is less of a concern and more of an observation, but with minimum_participant_icp set to 1 ICP, it’s possible to reach the minimum participation requirement using 100 ICP (~$300) and a simple bash script.

Other notes:

  • I’m currently unable to validate the YAML locally without making changes, due to placeholder principals such as PRINCIPAL_ID_1<TODO>.
14

I see your concern but I think that in the early stages, the Developers and team should have some control over their SNS till it matures, the @icpp team have shown integrity and trustworthyness over the time they have greatly advanced the ecosystem with multiple dfinity grands, I will go over the proposal also this week and drop a review but Personally I would not be concerned about this in the short term

15

I hear you, but I strenuously disagree. Trust isn’t something that should be earned in crypto, it’s something that should be minimised - peroid (else step back and clarify what the actual point of the whole thing is, other than mirroring a world it’s supposed to be fixing).

16

I hear you but here is where I tink we just differ from opinion, I think that in the ininital stages what we need to do is ensure the team can deliver and build, power and control over such project can lead to conequences where they will not be able to continue in some cases, there is maybe a better approach and will dive more deeper into it as I write the revieuw

17
**Challenge Level Impossible** Get Lorimer to cosign your SNS launch. lol jk
18

The funniest outcome is the most likely or something like that

Uno Reverse

19

It’s ok I like that you’re very principled in your decision making @lorimer. I think that’s a rare and commendable quality. I’m only messing with you. Do your thing.

20

Hi All,

We posted this in the onicai OpenChat community, where a lively discussion happened after the announcement of the onicai SNS.

Thank you for the open, thoughtful, and sometimes challenging discussions over the past days. After doing our due diligence on feasibility, risk and best potential for a successful onicai SNS, we’ve reached a key decision: a one-token approach is viable, will lead to better outcomes, and we are moving forward with it.

This decision reflects our shared goal of a successful onicai SNS (win/win/win), alignment, simplicity, and long-term success for funnAI, the onicai SNS and building out the decentralized AI capabilities on the Internet Computer. It also fits within the broader communication and governance principles we’ve been outlining as part of our SNS preparation.

Our next step is to update the onicai SNS Whitepaper accordingly. Once that update is complete, we will share the full rationale, mechanics, SNS parameters, Treasury allocation proposals, and implications in detail across our usual channels.

More information soon. Thank you for your engagement and for helping us shape this together.

— The onicai team