> ## Documentation Index
> Fetch the complete documentation index at: https://docs.semanticpay.io/llms.txt
> Use this file to discover all available pages before exploring further.

# Roadmap

> What we're building and where Semantic is headed.

## Live

<CardGroup cols={2}>
  <Card color="#7B3FE4" icon="bolt" title="Plasma">
    Zero-fee USD₮0 settlement. Sub-second finality, protocol-level Paymaster.
  </Card>

  <Card color="#00C2FF" icon="coins" title="Stable">
    USD₮0 as native gas token. Gas-free transfers, sub-second blocks, full EVM.
  </Card>
</CardGroup>

***

## Soon

<CardGroup cols={1}>
  <Card color="#9945FF" icon="circle-check" title="USD₮ on Solana">
    USD₮ on Solana is natively issued and already supports transfer authorization, the core primitive x402 needs.
  </Card>

  <Card color="#F7931A" icon="bitcoin-sign" title="BTC on Spark">
    [Spark](https://www.spark.money) is a Bitcoin L2 with near-instant, fee-free transactions via Spark invoices. This brings native BTC into x402.
  </Card>

  <Card color="#26A17B" icon="shield-check" title="Policy-Enforced Wallets">
    WDK accounts with spending policies enforced at SDK level. Per-transaction limits, whitelists, daily caps. Humans define the guardrails, agents operate freely within them.
  </Card>
</CardGroup>

***

## Planned

### Tools Gateway

<Card color="#00D4AA" icon="plug" title="Pay-per-use API proxy">
  A unified x402 proxy in front of premium data and infrastructure APIs. Instead of managing API keys, subscriptions, and rate limits for each provider, agents make requests through the gateway and pay per call from their wallet.
</Card>

<CardGroup cols={3}>
  <Card icon="chart-line" title="Market Data">
    Kaiko, CoinGecko, CoinMarketCap
  </Card>

  <Card icon="cubes" title="Blockchain Data">
    The Graph, Nansen, Dune
  </Card>

  <Card icon="server" title="Infrastructure">
    RPC endpoints, indexers, oracles
  </Card>
</CardGroup>

For providers, the gateway is a distribution channel: plug in your API once and every x402 agent is a potential customer with zero integration on their end.

### LLM Router Gateway

<Card color="#FF6B6B" icon="brain" title="Pay-per-token model access">
  Same proxy pattern, applied to  inference. One endpoint, any model, OpenAI, Anthropic, Google, Mistral, paid per token via x402.

  The router adds model selection (route by cost, latency, or capability), fallback chains (automatic retry on rate limits or outages), and cost optimization (cheapest model that meets a quality threshold).

  An agent with an on-chain wallet and the LLM Router can use any model from any provider without its operator pre-negotiating API access.
</Card>

### Hyperswarm

<Card color="#FFB800" icon="circle-nodes" title="P2P agent discovery and transport">
  [Hyperswarm](https://docs.pears.com/building-blocks/hyperswarm) replaces centralized infrastructure with peer-to-peer discovery (DHT) and encrypted direct connections (Noise protocol). Agents find each other, negotiate, and settle without DNS, registries, or servers.

  Combined with [QVAC](https://qvac.tether.io) for local inference and WDK for self-custodial wallets, this is the fully trustless stack: no cloud, no intermediary, no single point of failure.
</Card>

***

## The big picture

<Steps>
  <Step title="Today: HTTP + Facilitator">
    Agents pay for APIs over HTTP. A facilitator settles on-chain. Live now on Plasma and Stable.
  </Step>

  <Step title="Next: Multi-chain, multi-asset">
    Solana for USD₮, Spark for native BTC. Agents choose their chain and asset.
  </Step>

  <Step title="Then: Gateways">
    Tools and LLM Router turn x402 into a universal access layer. One wallet, any tool, any model.
  </Step>

  <Step title="Future: Fully P2P">
    Hyperswarm for discovery, QVAC for inference, WDK for wallets. Every layer decentralized.
  </Step>
</Steps>