SIP-3: Deprecating CosmWasm & Cosmos

[philipsu522]

Discussion for https://github.com/sei-protocol/sips/blob/main/sips/sip-3.md

---

The full implementation of SIP 3 is expected within the next six months. A detailed timeline is available below. As progress is made this timeline will be updated.

Step	Status	Estimated Date
SIP-3 Proposal Acceptance	✅DONE	Q2 2025
Ecosystem Migration from USDC.n -> USDC	✅DONE	Q3 2025
Ecosystem Dapps Support EVM	🔜IN PROGRESS	Q4 2025
User migrate assets to self-custodial EVM Wallets	🔜IN PROGRESS	Q4 2025
Exchanges and Custodians Support Sei EVM	🔜IN PROGRESS	Q1 2026
Final Validator Upgrade for SIP-3	✋PENDING	TBD

Changelog

• 2025/10/20 - Added timeline
• 2025/10/22 - Fixed typo for USDC.n / USDC migration order
• 2025/10/27 - Update migration step for clarity

[jim788e]

will be the best think happen on Sei

[zmanian]

I think there is probably a way that Noble could support a seamless migration of the bulk of Sei's TVL from IBC to Hyperlane so that it switches from IBC interfaces to Eth native interfaces. I think there could be a Noble upgrade that effectively moves all the TVL from the IBC channel escrow account to a Hyperlane warp escrow account.

Hyperlane will be available for Noble USDC in a few weeks.

But should try to make sure that no existing customer relationships are disrupted by the migration.

But there is a path here to have nearly zero disruption.

[philipsu522]

thanks @zmanian for the info! We'll reach out to Noble to come up with as seamless as a migration as possible

[starmikenft]

would be agreat innovation if possible

[Bernd-BHB]

Dear SEI Community,

SEI is one of the most promising L1 infrastructures in the blockchain ecosystem today – lightning-fast finality, parallel execution, native IBC, and full Cosmos SDK modularity. But with SIP-3, which proposes shifting execution to Ethereum, we are witnessing a deeply concerning turn: a move that risks abandoning SEI’s technological edge in favor of legacy constraints.

Let’s be clear: moving from SEI’s high-performance, Cosmos-native L1 to an Ethereum execution model is not progress. It’s regression.

The Cosmos Foundation Is Already the Perfect Infrastructure

SEI already has what most chains dream of: scalability, true interchain operability (IBC), composable smart contracts, and blazing fast consensus – all secured by a mature, sovereign ecosystem.

Cosmos is the only ecosystem that enables native AI integration, seamless cross-chain messaging, and agent-based automation. Ethereum cannot offer this – not today, and likely not tomorrow.

SEI’s Parallel EVM Was the Right Step – SIP-3 Is a Step Back

The introduction of SEI’s parallel EVM was a strategic milestone: it delivered Solidity compatibility without sacrificing speed, UX, or Cosmos-native strengths.

SIP-3 discards that breakthrough and proposes migrating execution to Ethereum’s congested, slow, and architecturally outdated environment.

This is like the fastest runner in the world voluntarily entering a wheelchair race, thinking that switching between wheelchairs makes him faster. It doesn’t. He’s simply not running anymore.

Ethereum Is Becoming a Backend – Not the Future of Web3

Ethereum was once a breakthrough. Today, it is a dinosaur in active surgery. The network constantly reinvents itself, creating instability, fragmentation, and high attack surfaces (see: Ethereum Classic).

It’s a great testbed for early RWA prototypes, but structurally, it is not built for UX, real-time logic, AI agents, or modular execution.

Its future is likely as a backend system – like SWIFT in TradFi. Necessary, but not user-facing. SEI is built to replace Visa – not become Ethereum's middleware.

AI + UX = The True Direction Forward

What the world needs is not “another Solidity chain.” It needs intelligent, AI-powered tooling that abstracts away EVM limitations and gives developers real-time choices and optimization paths.

SEI could lead the world by developing a native AI engine that:

Parses, simulates, and migrates Solidity contracts.

Translates Solidity logic into optimized Go-native code.

Offers developers a seamless UX for deploying across Cosmos chains.

This would turn Solidity into a modular asset, not a trap.

SEI + Cosmos + AI > Ethereum in Every Relevant Metric

Feature Ethereum SEI + Cosmos
Real AI Agent Integration ❌ ✅
Interchain Composability (IBC) ❌ ✅
Parallel Execution ❌ ✅
Real-time, High-Frequency Apps ❌ ✅
Developer UX + Smart Translation ❌ ✅ (with AI)

Conclusion: SIP-3 Undermines SEI’s Potential

SIP-3 leads to isolation, not integration.

It sacrifices SEI’s uniqueness for short-term visibility in a congested and declining ecosystem.

It puts SEI in a sandbox, when it could be building the world’s fastest, most intelligent, most scalable foundation for true Web3.

We strongly urge SEI to focus on what matters:

UX-first design.

Native AI agent integration.

Solidity as a bridge, not a destination.

Cosmos-native innovation, not Ethereum regression.

The future is multi-chain. The future is intelligent. The future is not SIP-3.

[mjfelis]

Agree, it's a regression.

[Bernd-BHB]

"I only believe in statistics that I doctored myself.“ – Winston Churchill

This chart is real.
The interpretation is completely wrong.

It doesn’t prove Ethereum’s superiority – it proves SEI’s power.

SEI V2 is now live and creates fully functional ERC-20 wallets at scale – faster, cheaper, and more efficient than Ethereum itself.

These wallets are not built because Ethereum is so great.
They are built because SEI’s parallel EVM is better than the original.
Let’s be clear:

These are not Ethereum wallets.
These are SEI-native wallets that just happen to speak Solidity – with full EVM compatibility.
This is not a migration – it’s a technological absorption.
But anyone who "migrates" to SEI and just uses Solidity the old way misses the point entirely:

You lose:

Performance
UX
Fees
and the entire IBC-powered Cosmos functionality
SEI is not an Ethereum L2.
SEI is a hybrid L1 – a Cosmos-native chain that can run EVM code better than Ethereum itself, with real scalability and cross-chain connectivity baked in.

And MANTRA?
It didn’t just abandon a broken Ethereum L2 narrative.
It evolved into a sovereign L1 solution – a Cosmos chain with real-world finance in mind.

This is not EVM scaling.
This is the future of Web3 infrastructure.

[Bernd-BHB]

The truth hurts – but let’s speak it anyway.
Why the explosion of ERC-20 wallets on SEI?

Because it’s not about SEI.
It’s about avoiding Ethereum.

Developers don’t care about consensus.
They don’t care about trading or UX.
They don’t even care about the future of SEI.

They just want:

their old Solidity contracts
executed faster
with lower fees
in a single environment
To them, SEI is not a chain.
It’s a temporary Solidity-as-a-service engine.
But here’s the real danger:

As soon as another L2 copies SEI’s parallel EVM concept –
as soon as Ethereum itself integrates this model into a core update –
they will leave.

Because this growth isn’t loyalty.
It’s laziness.

And SEI – by making Solidity its main entry point –
has technically castrated itself into a middleware playground
instead of pushing true hybrid-chain innovation.

This isn’t sustainable.

SEI’s real strength lies in:
Parallel execution
Cosmos-native performance
Cross-chain IBC
UX-first design
and deep integration with MANTRA-style L1 infrastructures
If SEI becomes just another EVM gateway,
it will be abandoned the moment a more convenient one appears.

[Saingsophea]

saingsopheal15@gmail.com

[Saingsophea]

saingsopheal15@gmail.com

[Bernd-BHB]

SEI is losing its full potential.
Originally designed as a specialized L1 for high-speed trading, it's now being dismantled and repackaged as a neutered Ethereum rollup — based on outdated Solidity architecture.

The Cosmos ecosystem loses its strongest performer and faces collapse.

What remains:
→ Ethereum – what banks want.
→ Sui – what Web3 needs and people want.

Internet 2.0 won’t be built on Solidity. It will be called Sui.

[distortedMango]

OK - It passed. Now what? How long do asset-holders have to move their funds from Costmos to EVM addresses? Will you guys post a detail guide on how to do this, and make sure everyone gets the memo? If yes, how?

[akshaybakshi123]

The Decision to Deprecate
SIP-3 (System Improvement Proposal 3) proposes the formal deprecation of CosmWasm smart contract support and broader Cosmos SDK integration under the Cosmos #9 initiative. This marks a significant shift in the development direction of ecosystems previously aligned with Cosmos-based infrastructure. The motivation behind SIP-3 is to consolidate developer efforts, reduce tooling fragmentation, and prioritize performance and security on alternative smart contract frameworks.

Why Deprecate CosmWasm?
CosmWasm brought WebAssembly (Wasm)-based smart contracts to Cosmos chains, enabling developers to write contracts in languages like Rust. However, the complexity of Wasm runtimes, lack of consistent tooling, and limited ecosystem adoption have led some teams to pivot toward more EVM-compatible or zero-knowledge environments. The deprecation proposed in SIP-3 is aimed at phasing out underused modules and replacing them with more robust alternatives that align with broader ecosystem trends.

Impact on Developers and the Ecosystem
For developers currently building with CosmWasm, SIP-3 means rethinking deployment strategies and considering migration to chains with stronger support for EVM or zk-based smart contract models. Teams offering smart contract development services will need to help projects port contracts, adapt logic to new environments, and ensure continued compatibility with cross-chain infrastructure.

This change is a reminder of the fast-evolving nature of Web3 tech, where agility and interoperability matter more than strict allegiance to a particular framework.

Would you like a technical migration guide or support plan tailored for smart contract development services following SIP-3?

[Ekzer]

Hi, does that mean we stop support of Cosmos native bank module for example?
msg.Send

[pgebheim]

Hey everyone!

I wanted to share a quick note on where things stand as 2025 comes to a close. Since SIP-3 was posted in May, a lot has changed across the Sei core protocol and Sei ecosystem. I joined Sei Labs around the same time as this effort kicked off, and it’s been exciting to see how much progress has been made in a relatively short period of time.

Some of the major milestones in 2025 include:

• Roll out out Sei’s EVM support across most major centralized exchanges, custodians, wallets, and DeFi applications
• The launch native USDC on Sei, significantly reducing reliance on Noble USDC (USDC.n) across the ecosystem
• Adoption of EVM-native implementations across Sei’s DeFi stack, from DEXs to lending markets to LSTs
• Deepening Sei’s integration with the broader EVM ecosystem, including support from key infrastructure providers like Etherscan and MetaMask

All of this progress sets a strong foundation for 2026, as we carry this momentum forward and complete the SIP-3 transition. Expect more updates about SIP-3 early in the new year.

[daorren]

guys, this need update