Router One

Web3 AI infrastructure

Crypto-Friendly LLM API Gateway for Web3 Developers

Router One gives Web3 teams one OpenAI-compatible endpoint for GPT, Claude, Gemini, DeepSeek, and more. Pay with USDT/USDC, set budget caps before agents spend, and trace every token, model, provider, latency, and cost.

Base URL
https://api.router.one/v1
API surface
OpenAI-compatible Chat Completions
Payments
USDT/USDC, Stripe, WeChat Pay, Alipay
Controls
API keys, budget caps, usage traces

Why Web3 teams use Router One

Stablecoin-friendly billing

Top up with USDT or USDC on supported networks when your team already manages treasury in stablecoins.

One API key for major LLMs

Route GPT, Claude, Gemini, DeepSeek, Mistral, and Llama through one OpenAI-compatible gateway instead of juggling provider accounts.

Budget caps for agents and bots

Create separate API keys for agents, scripts, and teams, then put hard spending limits around each workflow.

Cost and latency observability

Every request records model, provider, token counts, status, latency, and cost so agent spend is visible instead of surprising.

Web3 use cases

Router One is the LLM inference layer. Use Moralis, The Graph, Dune, Alchemy, QuickNode, Etherscan, or your own indexer for onchain data.

Onchain analyst agents

Summarize wallet activity, protocol usage, governance events, and risk signals from your own indexed data.

DAO and community assistants

Turn proposals, forum threads, Discord logs, and treasury updates into concise briefings for contributors.

Smart contract coding workflows

Use Claude Code or Codex CLI through Router One while reviewing Solidity, tests, audits, and deployment scripts.

Telegram and Discord research bots

Build assistants that answer market and protocol questions without giving financial advice or promising trading returns.

Trust boundaries

Not an agent runtime

Router One does not orchestrate tools, wallets, workflows, or blockchain transactions. Your app or framework owns agent logic.

No prompt or completion storage

Request and response bodies are processed in real time and not retained. Billing and observability use metadata only.

Transparent wallet billing

Wallet balance is denominated in USD after top-up. Model usage deducts at posted rates from the models page.

Compliance-aware wording

Crypto billing is for teams that prefer stablecoin workflows. It is not permission to bypass laws, sanctions, or upstream terms.

OpenAI SDK quickstart

Swap the base URL and keep your existing OpenAI-compatible client. Agent frameworks can use the same endpoint with per-agent API keys.

web3-agent.ts
import OpenAI from "openai";


const client = new OpenAI({
  baseURL: "https://api.router.one/v1",
  apiKey: process.env.ROUTER_ONE_API_KEY,
});


const result = await client.chat.completions.create({
  model: "auto",
  messages: [{ role: "user", content: "Summarize this wallet activity." }],
});

FAQ

Can I pay for LLM API credits with USDT?
Yes. Router One supports crypto top-ups with USDT on supported networks, credited into a USD wallet balance.
Can I pay with USDC?
Yes. Router One supports USDC top-ups on supported networks including Base and other listed chains.
Do AI agents get separate API keys?
Yes. Create separate API keys per agent, bot, team, or environment, then track and limit usage independently.
Can I set a hard budget cap before an agent spends?
Yes. Router One supports budget controls so a runaway agent or retry loop cannot spend indefinitely.
Does Router One store prompts or completions?
No. Router One does not retain request or response bodies. It stores metadata needed for billing, routing, and observability.
Is Router One an OpenRouter alternative for crypto teams?
Yes. Router One is OpenAI-compatible and adds stablecoin top-ups, China-friendly routing, budget caps, and request traces.

Related