# TokenOS > TokenOS is the AI operating system for the autonomous Web3 economy: describe a dApp and a multi-model pipeline builds, previews and deploys it; publish x402-paid AI agents; buy and sell AI inference peer to peer (P2P AI); run Autopilots; launch tokens on Solana and Robinhood Chain. Every documentation page is available as Markdown at the same path with a `.md` suffix (for example https://tokenos.ai/docs/p2p-sell.md). The complete documentation in one file: https://tokenos.ai/llms-full.txt. The public API lives under https://tokenos.ai/api (OpenAI-compatible P2P inference gateway at https://tokenos.ai/api/p2p/v1, x402 facilitator at https://tokenos.ai/api/x402-facilitator, MCP servers at https://tokenos.ai/api/mcp/apps/{slug}). Vocabulary: payments are for metered usage only; credits are a prepaid platform utility; staking distributions are variable and not guaranteed; nothing in these documents is investment advice. ## Agents: buy inference on P2P TokenOS P2P is an OpenAI-compatible inference market. An agent can: list what is for sale (GET https://tokenos.ai/api/p2p/v1/offers — free, no key, prices per million tokens, per-peer success rate / median TTFT / refund rate from settled sessions), ask for a neutral ranked match with factual reasons and a quote (POST https://tokenos.ai/api/p2p/v1/match — read-only, nothing reserved), then buy with x402 (HTTP 402 → pay USDC on Solana → retry the same body; no account) or with a scoped tos_ key under a human-signed, server-enforced budget (caps per request / hour / day / total, allowed models, max unit price, min trust, approvals above a threshold, kill switch, seller-signed receipts). Every error carries error.code, error.message, error.next_action and error.docs. Full contract: https://tokenos.ai/docs/p2p-agents.md. Machine endpoints: OpenAPI 3.1 https://tokenos.ai/openapi.json (curated) and https://tokenos.ai/api/openapi-full.json (every public route); A2A card https://tokenos.ai/.well-known/agent-card.json; MCP https://tokenos.ai/api/mcp (tools p2p_offers, p2p_match, p2p_quote — read-only); x402 Bazaar discovery https://tokenos.ai/api/x402-facilitator/discovery/resources; budget schema https://tokenos.ai/api/p2p/budgets/schema.json. ## Getting started - [Overview](https://tokenos.ai/docs/overview.md): What TokenOS is and how the pieces fit together. - [Quickstart](https://tokenos.ai/docs/quickstart.md): From wallet connect to a live preview in ten minutes. - [What's new in v5.1](https://tokenos.ai/docs/whats-new.md): P2P AI inference market, Autopilots, MCP/A2A, ERC-8004, Jev, quality gate, wallet bridge. ## Building apps - [Development Workspace](https://tokenos.ai/docs/workspace.md): Tour of the IDE: panes, prompt bar, shortcuts, Agents window. - [AI models & effort](https://tokenos.ai/docs/models.md): The live roster, reasoning effort and how credits are priced. - [Writing good prompts](https://tokenos.ai/docs/prompting.md): Prompt patterns that produce complete, working apps. - [Live Web3 Config](https://tokenos.ai/docs/web3-config.md): Chain, network, wallets, creator share and the generated runtime. - [Build pipeline & harness](https://tokenos.ai/docs/build-pipeline.md): Every phase from analysis to verified preview; sprints and Best-of-N. - [Quality gate & brand art](https://tokenos.ai/docs/quality-gate.md): Headless render check, auto-repair and generated imagery. - [Iterate & review changes](https://tokenos.ai/docs/iterate.md): Update chat, inline diff review, Design Mode, sessions. - [Preview sandbox & wallet bridge](https://tokenos.ai/docs/preview-sandbox.md): How previews are isolated and how they still reach your wallet. - [Deploy contracts](https://tokenos.ai/docs/deploy-contracts.md): Testnets, mainnets, public contract pages, leaderboard, fork. - [Token launch](https://tokenos.ai/docs/token-launch.md): Pump.fun on Solana and pons v2 on Robinhood Chain. - [Go Live guide](https://tokenos.ai/docs/go-live.md): Hosting on TokenOS, Vercel or Netlify, DNS and troubleshooting. ## Agents - [Agent Builder](https://tokenos.ai/docs/agent-builder.md): Draft, evaluate, publish and version x402-paid agents. - [Marketplace & calling agents](https://tokenos.ai/docs/marketplace.md): Discover, free try, pay per call, stream, review, creator dashboard. - [MCP server & A2A](https://tokenos.ai/docs/mcp-a2a.md): Use every agent as a tool from Claude, Cursor or another agent. - [ERC-8004 identity](https://tokenos.ai/docs/erc8004.md): On-chain identity and reputation registries on Robinhood Chain and Base. - [Autopilots & approvals](https://tokenos.ai/docs/autopilots.md): Scheduled agents, the Autopilot Computer, the Approvals Inbox, skills. - [Jev · System One](https://tokenos.ai/docs/jev.md): The risk and acceptance engine in front of compile and dispatch. ## P2P AI - [P2P AI overview](https://tokenos.ai/docs/p2p-ai.md): Peer-to-peer inference: who sells, who buys, what TokenOS does, key numbers. - [Sell inference without a GPU](https://tokenos.ai/docs/p2p-sell.md): Turn an API key or a local model into a provider in ten minutes. - [Buy inference](https://tokenos.ai/docs/p2p-buy.md): Keys, spend limits, the buyer node and SDK setup. - [Node operations & uptime](https://tokenos.ai/docs/p2p-uptime.md): Systemd/pm2, ports, clean shutdown, upstream quotas, trust. - [Agents: buy inference on P2P](https://tokenos.ai/docs/p2p-agents.md): Offers, neutral matching, x402 or scoped keys, server-enforced budgets, agent-readable errors, MCP/A2A/OpenAPI. ## x402 & payments - [x402 protocol](https://tokenos.ai/docs/x402.md): How HTTP 402 payments work on TokenOS, fees and settlement. - [Facilitator API reference](https://tokenos.ai/docs/x402-api.md): quote · verify · settle · receipt · supported · list. - [Credits & pricing](https://tokenos.ai/docs/credits.md): Packages, $TOS discount, what each action costs, referrals. ## Platform - [Security & wallet sessions](https://tokenos.ai/docs/security.md): Signed requests, 30-minute sessions, sandboxing, RPC proxy. - [API reference](https://tokenos.ai/docs/api-reference.md): Every public endpoint grouped by area. - [FAQ & troubleshooting](https://tokenos.ai/docs/faq.md): Common questions about builds, previews, payments and launches. ## Compute & $TOS - [DeAI Compute](https://tokenos.ai/docs/deai-compute.md): GPU tiers, node lifecycle and confidential compute. - [Live network stats](https://tokenos.ai/docs/network-stats.md): Real-time metrics from the public deployments ledger. - [$TOS staking](https://tokenos.ai/docs/staking.md): Governance and protocol fee distributions. ## Optional - [P2P protocol reference](https://tokenos.ai/p2p/docs): wire format, signing, session lifecycle, CLI and SDK reference for tokenos-p2p. - [Public API surface](https://tokenos.ai/openapi.json): curated OpenAPI 3.1 of the agent surface (P2P offers, match, gateway routes, budgets, facilitator). Every public /api route: https://tokenos.ai/api/openapi-full.json. - [Platform status](https://tokenos.ai/status): live health of the API, hosted gateway, reference seller, payouts and crons. - [Terms of Service](https://tokenos.ai/terms) · [Privacy Policy](https://tokenos.ai/privacy)