H1 2026 0G Ecosystem Update

Six months of building on the largest AI Layer 1. 400+ integrations, 350+ projects, and AI is now the leading and fastest-growing vertical, verifiable by design.

0G began as a bold vision for open, verifiable AI infrastructure, and it has since become the largest and fastest-growing AI Layer 1. Our H2 2025 update told the story of mainnet, the exchanges, and the validators that came with it. This update tells a different story: what builders are actually building on 0G, and AI is the leading and fastest-growing vertical.

Since our last ecosystem report, the network has grown to 400+ integrations across 350+ projects. The shape of that growth matters more than the count. The projects shipping on 0G today are autonomous agents, frontier models, verifiable media-safety tools, AI-native exchanges, and private personal AI. More and more of them are the kind of enterprise-grade, verifiable, and private AI that needs infrastructure it can prove rather than trust.

As in every prior report, the total number of integrations exceeds the number of individual projects building on 0G, because a single project often builds across more than one 0G layer. That composability is the point. A team can store training data, run inference, and settle agent actions on the same stack, and many of the projects below do exactly that.

0G Adoption by Layer: Where Builders are Integrating

0G is the trust layer for AI: one stack that runs from raw hardware to the apps people use, built so AI work can be verified rather than trusted. Three layers sit on a Physical Infrastructure base, and the clearest read on where the AI build-out is happening is which parts of the stack builders reach for.

AI Private Cloud. The foundation that builders integrate first, spanning infrastructure, execution, and developer tooling. This is where 0G Chain, 0G Storage, and 0G Compute live, and it carries the bulk of the ecosystem's integrations today.

• 0G Chain (200+ integrations). Settlement and coordination for onchain AI. Agents register identity, anchor proofs, and settle actions here, from DeFi and gaming to the coordination autonomous agents need to act with finality.
• 0G Storage (90+ integrations). The onchain home for AI data: training data, datasets, model checkpoints, and agent memory. It is what lets an agent remember, a model reproduce, and a dataset stay verifiable long after it was created.
• 0G Compute (60+ integrations). Verifiable AI inference, run inside trusted execution environments. This is the layer powering the model and agent stories below, where prompts become outputs a user can independently check.

Storage and compute are the layers AI builders depend on most directly, and that tilt anchors the rest of this report.

Agentic Alignment. The layer above the cloud, where models, alignment, agent harnesses, and onchain finance come together so autonomous software can act safely and with economic context. This is where the agent economy is taking shape, as software moves from demos to production.

Apps. The top layer, where the stack reaches users: the ecosystem of products, operations, physical AI, and agents people actually run. The case studies in the next section live here.

All of it rests on Physical Infrastructure, the hardware and networking underneath the verifiable stack.

Who's Building with 0G: Our Expanding DeAI Ecosystem

The logo wall is the breadth: every qualified project building on 0G, grouped by layer and vertical. This section is the depth. Rather than list every name, we picked the strongest AI case studies in each category, the projects that show what the stack actually enables.

AI Agents

The deepest pool on 0G, and the clearest signal of where the network is heading. These are not chatbots. They are agents that hold identity, memory, and the ability to act onchain.

• Ghast AI is a browser-native AI assistant that lives in a Chrome extension, runs its inference on 0G Compute, and checkpoints its memory onchain to 0G Storage under the ERC-7857 Agentic ID standard, so the intelligence is something the user truly owns and can carry across agents rather than rent from a platform. It is the clearest end-to-end example of the full 0G stack in one product. Read the Ghast AI deep-dive.
• Dolly turns community activity into ownable, onchain identity and engagement assets minted as Agentic IDs, building on 0G through the AIverse marketplace.
• Neosoul is a broad AI platform spanning agents, apps, and data, building on 0G Storage and the Agentic ID standard.
• Clawdspiracy is an AI-agent game built on 0G, with its agents running on 0G Compute.

AI Apps

Consumer and prosumer AI products that chose 0G for the parts that have to be private or verifiable.

• 0G App is 0G's own app at app.0g.ai. Through 0G Studio, anyone can turn a plain-language prompt into a deployable web app from the browser, with inference running on 0G Compute inside trusted execution environments and work syncing to 0G Storage for verifiable, password-free restore.
• ClaraVerse is an open-source, privacy-first AI workspace for chat, agents, and retrieval, and it uses 0G Compute as a cloud inference provider through its public 0G Compute adapter.
• Flashback Labs builds privacy-first personal AI, "AI twins" assembled from your own memories, and uses 0G Chain, Storage, and Compute as part of its stack.
• Kodeus AI is an agentic operating system with onchain agents, building on 0G Compute and Storage.
• Matterhorn is an AI-native development tool that turns natural-language prompts into deployed smart contracts and Web3 apps, hosting and routing its models on 0G Compute and using 0G Storage and Chain in its stack.
• Arc Terminal, from Kaiko Studios, is an early-stage AI operating-system platform building across 0G Chain, Compute, and Storage.
• Kult.Games is an AI-native gaming app on 0G, where play-and-earn titles run their game logic onchain on 0G Chain.

Privacy and Verifiable AI

The enterprise hook. When the data is sensitive or the output has consequences, the AI has to be provable. These teams build on 0G specifically so they can prove what their AI did.

• 0G Agentic ID is 0G's own onchain identity and verifiable-behavior layer for AI agents. Built on the ERC-7857 Agentic ID standard and aligned with the ERC-8004 trustless-agent standard, it gives each agent a self-owned identity whose signing and data keys live inside its own trusted execution environment, so the identity cannot be impersonated or seized by any platform. Every agent response carries a TEE-signed serve-proof, so anyone can cryptographically verify a response truly came from the attested agent, which keeps reputation sybil-resistant. It is live on 0G testnet at agenticid.0g.ai and rolling out as 0G's emerging agent-identity layer.
• Veyra is a verifiable AI video-safety agent that scores deepfakes and harmful content, running its analysis on 0G Compute and anchoring each safety decision as a cryptographic proof on 0G.
• Morphic is an AI co-host for short-term rental hosts that runs its models inside trusted execution environments on 0G Compute, so each output can be verified rather than trusted.
• Ghast AI belongs here too. Every action it takes is checkpointed onchain, which is what turns a browser agent into one a user can actually audit.

AI Models and Training

The supply side of intelligence. Frontier models and model infrastructure are arriving on 0G, which means agents can call them with the same verifiability as everything else on the stack.

• Alibaba Cloud brings the Qwen model family to onchain AI agents, with 0G providing verifiable, token-gated access to inference on Alibaba Cloud Model Studio. See the deep-dives on the Alibaba Cloud x Qwen integration and on Qwen 3.7 Max live on 0G.
• MiniMax is the frontier lab behind M3, which debuted as the number one open-weight model on Artificial Analysis and a top trending model on OpenRouter. M3 is live on 0G Compute. Try M3 on 0G.
• 0G Private Computer is 0G's own product for private, decentralized model hosting, where large language models like MiniMax M3 run inside trusted execution environments and users can verify both the model and its output.

DeFAI

AI applied to onchain finance, where the agents place the trades themselves.

• Zia Finance / TradeGPT is a natural-language DEX native to 0G, where TradeGPT turns plain-English prompts into trades and an AI liquidity layer rebalances across its markets with just-in-time provisioning.
• Envy Trade brings AI to onchain trading, using 0G Storage as its data layer.
• Tempora Labs builds DR HIRO, an agentic DeFi portfolio copilot aimed at institutional funds, using 0G Storage as its memory and state layer.

DeFi, Infrastructure, and Payments

The onchain economy the agents plug into: exchanges, settlement, staking, and the payment rails that move value between them.

• Oku brings production DeFi to 0G, with Uniswap V3 liquidity and Morpho vaults live on the network alongside an active incentives campaign with UniV4 on the way.
• Voyager is an institutional multi-asset exchange building on 0G Chain.
• Khalani Network builds intent-based cross-chain payments with fiat and crypto on-ramps, settling on 0G Chain.
• American Fortress is a privacy and payments layer with "Send-to-Name" stealth addresses for people and AI agents, and announced a partnership to co-build private AI-agent trading infrastructure with 0G.
• Pi Squared is a zero-knowledge settlement layer building its h402 payment rail on 0G.
• Gimo Finance brings liquid staking and restaking to 0G Chain.
• Neuron.world connects IoT devices to onchain data and is live on 0G mainnet storage.
• Phala Network, the decentralized confidential-compute cloud with TEEs, announced a partnership to bring its compute to 0G.

The Road Ahead

Two reports ago, the question was whether decentralized AI infrastructure would find real builders. That question is settled. The story now is the quality of what gets built: agents that hold identity, models that arrive with proofs, verifiable AI for when the output has consequences, and finance that an agent can run end to end.

The next half pushes further in the same direction. More frontier models on 0G Compute. More enterprise teams choosing verifiable and private AI over the centralized default. And a maturing agent stack, anchored by the ERC-7857 Agentic ID standard, that gives autonomous software a real identity to act under.

Building AI that has to be verifiable or private? Start at build.0g.ai, or reach the integrations team at 0g.ai/partners.

Frequently asked questions

What is the H1 2026 0G ecosystem update?

It is the latest entry in 0G's half-yearly ecosystem report, covering the projects building on 0G in the first half of 2026. This update spans 400+ integrations across 350+ projects, with the spotlights focused on the strongest AI case studies in each category.

How does 0G support AI builders?

0G is built as a modular AI stack, spanning AI Private Cloud, Agentic Alignment, and Apps on a Physical Infrastructure base. Builders integrate against its core primitives inside the AI Private Cloud: 0G Chain for settlement and onchain coordination, 0G Storage for AI data, datasets, and agent memory, and 0G Compute for inference inside trusted execution environments. A single project can build across all of them.

What is 0G?

0G is the blockchain for AI. It combines chain, storage, and compute in one stack where AI work, from training data to inference to agent actions, can be verified onchain. The Aristotle mainnet has been live since September 2025.

Where can I get 0G?

get.0g.ai is the interactive guide on how and where to acquire 0G tokens, with options for developers, investors, and ecosystem participants.

Where can I build on 0G?

Start at build.0g.ai for documentation, starter kits, and the latest SDK. Visit 0g.ai/partners to browse the projects already shipping on the network.

Stay connected

Follow the build. X (@0G_labs) | Website | Blog | Partners | Build

Sources:

• 0G H2 2025 Ecosystem Update
• 0G H1 2025 Ecosystem Update
• 0G Partners directory
• Ghast AI on 0G
• Alibaba Cloud x Qwen on 0G
• Qwen 3.7 Max live on 0G