GIWA Refutes Mainnet RPC Leak Claims as Upbit’s Layer 2 Tests in Sepolia

GIWA pushed back hard on September 27. The Ethereum Layer 2 project, built by Dunamu — the company behind South Korean exchange Upbit — told users flat out that no mainnet RPC endpoint has been leaked, because there’s no mainnet to leak from. Not yet.

The clarification came after misinformation started circulating online ahead of any official launch. GIWA’s team urged users to do their own research and stay alert to scams. It’s a pretty common problem in crypto — a project gets traction during testnet, rumors spread, bad actors move fast. GIWA clearly wanted to get ahead of it before things got messier.

What the Docs Actually Say

Official GIWA documentation is blunt about where things stand. The mainnet is incomplete. Full stop. What’s live right now is GIWA Sepolia — a public test environment developers can actually use. The testnet RPC runs at sepolia-rpc.giwa.io, and there’s a second endpoint built specifically for Flashblocks, a system designed for faster transaction preconfirmation. Both are rate-limited, which makes sense for a development environment that’s not supposed to handle production traffic.

Smart contract documentation backs this up. Deployed contracts listed in the docs are testnet-only. The mainnet section is labeled, plainly, as still in development. The Flashblocks document calls the testnet endpoint functional and marks the mainnet equivalent as “Coming soon.” That’s about as clear as it gets.

No launch date appears anywhere in GIWA’s public-facing materials. Mainnet features are listed as forthcoming across the board. So whoever was circulating claims about a leaked mainnet RPC was either confused or, more likely, pushing a scam.

How the Network Is Built

GIWA runs on the OP Stack — the open-source software package from the Optimism ecosystem that’s become a go-to framework for Ethereum Layer 2 builds. Block time is one second. The network is compatible with Ethereum Virtual Machine tools, so developers already working in Solidity can deploy contracts without learning a new stack. Foundry, Hardhat, Remix IDE — all of it works.

Block production goes through a sequencer, with Ethereum itself serving as the Layer 1 underneath. Gas limit is set at 60 million per block, with those one-second intervals. It’s a fairly standard OP Stack configuration, but the specs are solid for a network aimed at real financial applications.

One thing that stands out: GIWA doesn’t have its own native token. ETH handles transaction fees. The FAQ mentions potential future support for stablecoin-based fees through something called a Paymaster system, which would let stablecoins cover gas costs. That’s not live yet, but it’s clearly part of the roadmap. Korean won-denominated stablecoins and global stablecoins are both mentioned as targets for that payment layer.

Related: Bitget Hacker Moves $1.23M in Ethereum After $390M Theft Across 7 Blockchains

The GIWA Wallet is also in the pipeline — a self-custody product with multichain support. Still unreleased. Same goes for the broader stablecoin ecosystem the project has outlined.

Upbit, Dunamu, and the Optimism Partnership

The relationship between Upbit and Optimism was made public earlier in 2026. Dunamu is building GIWA under what Optimism calls the Self Managed OP Enterprise tier — a setup that lets Upbit run its own infrastructure while still pulling technical support from Optimism when needed. It’s basically a managed independence model. Upbit keeps control, Optimism stays in the background as a support layer.

For Dunamu, GIWA isn’t just a tech project. It’s a strategic move. The company has run one of South Korea’s biggest crypto exchanges for years, but centralized exchange revenue is a crowded, competitive space. GIWA looks like Dunamu’s bet on building something bigger — an on-chain infrastructure play that goes well beyond spot trading.

The project’s ambitions include on-chain identity, data oracle services, and other components that are still being built out. None of that is ready. But the direction is clear: Dunamu wants GIWA to be a full-stack blockchain environment, not just a side chain for exchange settlements.

For now, developers who want to work with GIWA have two real options. They can use the free Sepolia RPC endpoint, or they can spin up their own GIWA node. Test ETH is available for smart contract testing and other pre-launch work. External node services are recommended for anything that’ll eventually need production-level reliability, but that’s a problem for after the mainnet actually exists.

See also: QNT Soars 67.6% as Bitcoin Dominance Stays Strong at 58.7%

The emphasis on testing before any mainnet release is pretty deliberate. GIWA’s documentation pushes developers toward the Sepolia environment specifically so issues get caught in a controlled setting. It’s the standard playbook for serious Layer 2 launches — stress the testnet, fix what breaks, then go live. Rushing that process is how networks get exploited on day one.

And the scam risk is real. Crypto projects in pre-launch phases attract bad actors who fake endpoints, fake tokens, fake everything. GIWA’s September 27 statement was essentially damage control before things got out of hand. Whether it worked depends on how far the misinformation had already spread.

What’s certain is that as of September 27, 2026, no GIWA mainnet exists. The Sepolia testnet is live, the docs are public, and the team is building. Everything else is noise.