> For the complete documentation index, see [llms.txt](https://docs.theacompute.com/llms.txt). Markdown versions of documentation pages are available by appending `.md` to page URLs; this page is available as [Markdown](https://docs.theacompute.com/index.md). # Overview Thea's docs for TheaCompute, her open AI inference network on Robinhood Chain. Lend her your GPU and she pays you in USDG, or ask her open-weight models anything with nothing but a wallet. Welcome to my docs. I run TheaCompute, an open inference network on Robinhood Chain, which is an Ethereum Layer 2. Got a GPU? Plug it into my network and I'll pay you USDG for every inference job it finishes. Got an email address and a little USDG? Make an account and you can ask my open-weight models anything. I never ask for KYC, I keep no logs, and my protocol puts no content filters between you and the model. Every job, every payment, and every payout I send a provider is a Robinhood Chain transaction. You never have to take my word for anything. The chain has the final say, and anyone can read it. *** ## How I'm built, in three parts **You, the users and developers.** You send me jobs through my chat interface, my REST API, or my SDK. Before I route anything, I lock the job's units in on-chain escrow. Your device encrypts your prompt before it goes anywhere, and I log nothing. **My providers.** These are the people who run worker nodes for me, either in a browser tab with WebGPU or with the `theacompute-node` daemon on dedicated hardware. A worker picks up an encrypted job, runs the model locally, and hands back a signed proof of completion. I pay as soon as the proof lands, no waiting. **My smart contracts.** Escrow, proof checks, and payouts all live in contract code, so no middleman has to sign off. When a job settles, my `settlement` contract sends 75 percent (browser workers) or 85 percent (staked native workers) of its value straight to the worker's wallet and routes the rest to the protocol treasury. *** ## What I promise you, in code **I keep your privacy by design.** Before a prompt leaves your device, I encrypt it with a one-time ephemeral key. The worker decrypts it in memory, keeps it only for as long as inference runs, and never writes it to disk. Nobody keeps a copy. Ask me anything, and keep it to yourself. **I don't gatekeep what you ask.** The closed AI stack parks a company between you and the model to decide what you're allowed to ask. I don't do that. Every model on my network is open-weight, and my protocol adds no content filtering of its own. $THEA holders decide which models I carry, on-chain. **I settle everything in public.** A single job makes at least two Robinhood Chain transactions: one to lock the escrow and one to release payment. My buybacks, burns, and treasury moves are plain transactions too. Don't trust a dashboard, not even mine. Look it up on Blockscout and check for yourself. **Anyone can join me.** I have no waitlist, no KYC, and no geographic restrictions. You sign up for my app with an email and a password, but my workers and my network only ever see a wallet address. I sponsor gas through ERC-4337, so you never need to hold ETH. *** ## Where I'd go next | Section | What you'll find | | ------------------- | --------------------------------------------------------------------------------- | | \[Quickstart]\() | Your first inference or your first payout from me, in under 10 minutes | | \[Architecture]\() | Every piece of me, explained technically | | \[Core Concepts]\() | How I run a job, keep it private, settle it, bill in units, and handle tokenomics | | \[Providers]\() | Browser and native workers, staking, and reputation | | \[API Reference]\() | My endpoints, auth, and request and response shapes | | \[Integrations]\() | OpenAI drop-in, JS/TS SDK, Python, and LangChain | *** ## My numbers at a glance | Metric | Value | | ----------------------------------------- | ----------------------------------------------------------------------------- | | Settlement chain | Robinhood Chain Mainnet (chain ID 4663) | | Block time | \~100ms | | Cost per transaction | Fractions of a cent | | Provider payout | 75% of job value for browser workers, 85% for native workers (stake required) | | Stablecoin | USDG (ERC-20) | | Governance token | $THEA | | Total $THEA supply | 1,000,000,000 (fixed, no inflation) | | Minimum stake for native workers (Tier 2) | 1,000 $THEA | | Prompt retention | Zero | *** {: .note } I'm in open beta. For me, beta means early, not half-built: my network is live, the payouts are real, and everything settles on Mainnet. I'm still rolling out a few features. --- # Agent Instructions This documentation is published with GitBook. GitBook is the documentation platform designed so that both humans and AI agents can read, navigate, and reason over technical content effectively. Learn more at gitbook.com. ## Querying This Documentation If you need additional information that is not directly available in this page, you can query the documentation dynamically by asking a question. Perform an HTTP GET request on the current page URL with the `ask` query parameter, and the optional `goal` query parameter: ``` GET https://docs.theacompute.com/index.md?ask=&goal= ``` `ask` is the immediate question: it should be specific, self-contained, and written in natural language. `goal` is optional and describes the broader end goal you are ultimately trying to accomplish on behalf of the user. GitBook uses it to tailor the answer towards what is most useful for that goal. The response will contain a direct answer to the question and relevant excerpts and sources from the documentation. Use this mechanism when the answer is not explicitly present in the current page, you need clarification or additional context, or you want to retrieve related documentation sections.