Cointime

Download App
iOS & Android

IMO: Initial Model Offering

From ethresearch by fewwwww

TL;DR

0. IMO: Initial Model Offering

In this age of AI, we are introducing a new mechanism, IMO (Initial Model Offering):

  • IMO launches an ERC-20 token (more specifically, ERC-7641 Intrinsic RevShare Token) of any AI model to capture its long-term value.
  • Anyone who purchases the token becomes one of the owners of this AI model.
  • Token holders share revenue of the IMO AI model.

Many open-sourced AI models face the challenge in monetizing their contributions, leading to a lack of motivation for contributors and organizations alike. As a result, the AI industry is currently led by closed-source, for-profit companies. The winning formula for open-source AI models is the need to gather more funding and build in public.

With IMO, we can win the fight for open-source AI. IMO can enable the sustainability of the open-source AI model’s ecosystem by fostering long-term benefits and encouraging engagement and funding to the open-source AI community. The win is when we have better open-source models than proprietary models.

IMO tokenizes the ownership of open-source AI models, sharing its profit to the token holders.

1. IMO Mechanism

Anyone can own model and receive its revenue by simply holding ERC-7641 token.

In the current design, tokens from IMO will be issued on the Ethereum as ERC-20 and ERC-7641 token. Token holders will receive a share of profits from its model revenue and its generated NFT fee (eg. the royalty fee of AIGC-NFT generated from the Stable Diffusion model).

The overall architecture of the IMO mechanism is illustrated like this:

Untitled (4)1920×1080 506 KB

In IMO, there are two core components:

  • Onchain AI Model with Verifiability
  • Revenue Sharing of Onchain Usage

a) Onchain AI Model

We need a way to run IMO AI models fully on chain and verifiably. This is critical because we need to make sure that tokens are indeed tied to the right model and the right AI inference.

There are two mainstream approaches to implementing onchain AI models: zkML (by generating cryptographic proofs of the model) and opML (by using cryptoeconomics to ensure that the model is correct). We also invented opp/ai, a combination of two approaches with both of their advantages. Based on practicality and performance, we adopt opML to port AI models onchain for now.

opML is at the core of OAO (Onchain AI Oracle), which is essential to bring AI models to IMO.

b) Revenue Sharing through ERC-7641 Intrinsic RevShare Token

Holders of IMO tokens will receive the benefits of revenue streams including but not limited to:

  • Revenue of model usage (Model Ownership, ERC-7641 Intrinsic RevShare Token): Each use of the AI model onchain will incur a fee, which will be distributed to IMO tokens.
  • Revenue of AI-generated content (Inference Asset, eg. ERC-7007 Zero-Knowledge AI-Generated Content Token): Each use of the AI model generates a specific output and result (e.g. Stable Diffusion for an image NFT and Sora for a video NFT), which may carry a royalty fee and a mint fee that can be distributed to IMO tokens.

Untitled (5)1920×1080 67.3 KB

The AI model distributes the fees directly to the IMO tokens. We have introduced an extension to ERC-20: ERC-7641, which standardizes revenue-sharing logic.

Untitled (6)1920×1080 120 KB

Holding IMO tokens is owning a part of an AI model and its future revenue. Imagine now you own a part of LlaMA2, Stable Diffusion, or Mistral. This not only benefits token holders but also contributes to the sustainable funding of open-source AI models.

2. Future of AI x Crypto is IMO

IMO is the future of AI x Crypto. Own AI model onchain and earn its revenue with IMO. Stay tuned and more updates soon!

Watch our talk “AI Isn’t Evil and Smart Contract Proves It 3” at ETHDenver on IMO.

Dive deep into our works on Onchain AI:

Comments

All Comments

Recommended for you

  • Trump Shares Op-Ed Claiming He is Winning the War Against Iran

    On August 29, U.S. President Trump shared a commentary article from the New York Post on Truth Social on Saturday, which stated that he is winning the war against Iran and should maintain the current strategy. The title of the article Trump shared read: 'Trump is Winning the War Against Iran - Stay the Course.'

  • Morgan Stanley: 2028 as a Key Observation Point for Global Memory Competition Landscape

    On August 29, Morgan Stanley pointed out that the rise of Chinese memory manufacturers should not be viewed merely as a technological catch-up or low-cost substitution; what is truly noteworthy is that their production capacity may gradually become large enough to alter the supply structure of the global memory market. Changxin Technology and Yangtze Memory Technologies are currently entering the mainstream product market and gradually extending into high-profit markets such as HBM, high-end server DRAM, and enterprise SSDs. Morgan Stanley considers 2028 as an important observation point for the global memory competition landscape. From 2026 to 2027, demand for AI servers, capacity crowding of advanced wafers by HBM, import substitution, and the time required for customer certification may absorb most of the new supply from Chinese memory manufacturers. By 2028, as Chinese manufacturers expand production, the additional capacity from Samsung, SK Hynix, and Micron, which had previously initiated expansions, will also be released. At that time, the supply variables in the global memory market will significantly increase. Morgan Stanley estimates that Changxin's DRAM monthly production capacity will rise from 180,000 wafers in 2025 to 300,000 in 2026, accounting for approximately 13% of global DRAM wafer capacity and about 11% of bit shipments; by 2028, it is expected to further increase to 500,000 wafers, and by 2031, it could reach 800,000 wafers. If the expansion proceeds smoothly, Changxin's global DRAM bit shipment market share could approach 15% by 2030, and it may even have the opportunity to surpass Micron in production capacity around 2028, becoming the third-largest DRAM supplier in the world.

  • US Spot Ethereum ETF Sees Net Inflow of $102.17 Million Yesterday

    On August 29, according to monitoring by Trader T, the US spot Ethereum ETF recorded a net inflow of $102.17 million yesterday.

  • US Spot Ethereum ETF Sees Net Inflow of $102.17 Million

    On August 29, according to monitoring by Trader T, the US spot Ethereum ETF experienced a net inflow of $102.17 million yesterday.

  • US Spot Bitcoin ETF Sees Net Outflow of $201.81 Million

    On August 29, according to monitoring by Trader T, the US spot Bitcoin ETF experienced a net outflow of $201.81 million yesterday.

  • US Spot Bitcoin ETF Sees Net Outflow of $201.81 Million Yesterday

    On August 29, according to monitoring by Trader T, the US spot Bitcoin ETF experienced a net outflow of $201.81 million yesterday.

  • BTC Surpasses $78,000

    Market data shows that BTC has surpassed $78,000, currently priced at $78,009.49. The 24-hour decline has narrowed to 3.23%. Due to significant market fluctuations, please ensure proper risk management.

  • BTC Surpasses $78,000

    Market data shows that BTC has surpassed $78,000, currently priced at $78,009.49, with a 24-hour decline narrowing to 3.23%. The market is experiencing significant volatility, so please ensure proper risk management.

  • BTC Briefly Drops Below $77,000

    Market data shows that BTC briefly fell below $77,000, currently reported at $77,694, with a 24-hour decline of 3.3%. The market is experiencing significant volatility, so please ensure proper risk management.

  • BTC Briefly Drops Below $77,000

    Market data shows that BTC briefly fell below $77,000, currently reported at $77,694, with a 24-hour decline of 3.3%. The market is experiencing significant fluctuations, so please ensure proper risk management.