BridgeNode MCP

io.github.bridgenode-ai由 GitHub 账号 bridgenode-ai 发布v1.8.1更新于 Oct 24, 2026

BridgeNode — x402 pay-per-request AI inference. OpenAI-compatible API + MCP, Solana USDC, gas-free.

已验证Streamable HTTP可网页运行金融AI 与机器学习

概览

AI 生成的概览

BridgeNode MCP 让助手调用兼容 OpenAI 的多模型对话补全,按次以 Solana USDC 通过 x402 付费,并提供免费模型与试用额度。

可以做什么
  • 向支持的模型发送对话补全请求,支持 OpenAI 风格的 tools 与 tool_choice,并返回工具调用
  • 列出可用模型及其价格
  • 在付费前估算一次对话请求的 USDC 费用
  • 免费模型和每个客户端对付费模型的首次调用无需付款,之后的调用通过 x402 自动支付
适用场景
适合助手需要模型推理但不想注册账号或申请 API 密钥的场景,也适合希望按次付费并设置消费上限、而非订阅制的用户。
运行要求
通过 npx 以本地 stdio 进程运行(需 Node.js),并代理到远程 streamable-HTTP 端点,也可直接作为远程端点使用。无需 API 密钥或注册。付费调用需要 Solana 钱包私钥,放在 BRIDGENODE_WALLET_KEY 中,并可选用 BRIDGENODE_MAX_PER_CALL 与 BRIDGENODE_DAILY_CAP 设置消费上限。需要网络访问。
安装前请注意
付费调用会花费真实的 Solana USDC;包装器会用 BRIDGENODE_WALLET_KEY 中的钱包私钥自动签名付款,该私钥可支配资金,建议使用专用的低余额钱包。消费策略默认按失败即拒绝处理,但仍应主动设定。提示词与请求会发送给第三方 BridgeNode 服务。

安装

在 SourceWeft 中

  1. 打开 控制台中的 BridgeNode MCP,将其添加到工作区。
  2. 为需要使用其工具的对话启用该服务。

Web executable,通过 Streamable HTTP。 远程服务在工作区中配置后即可从网页运行时运行。

交给智能体安装

把这段提示词粘贴到 Claude Code、Codex、Cursor 等智能体,或 SourceWeft 对话中。智能体会读取该服务器的安装指南,先向你展示发布者、端点和登录方式,经你同意后再把它添加到你的客户端。密钥不会写进任何命令。

阅读 https://sourceweft.com/mcp/io.github.bridgenode-ai%2Fbridgenode-mcp/install.md,并按其中说明把这个 MCP 服务器接入我的客户端。修改任何配置之前,先告诉我它的发布者、端点和登录方式,等我确认。不要把 API 密钥写进命令或我会提交的文件;问我密钥应从哪里获取。

查看智能体所依据的安装指南

其他 MCP 客户端

把它添加到你客户端的 mcpServers 配置中。

{
  "mcpServers": {
    "bridgenode": {
      "type": "http",
      "url": "https://bridgenode.cc/mcp"
    }
  }
}

README

@bridgenode/mcp

[npm version] [npm downloads] [License: MIT-0] [CI] [Website] [BridgeNode on x402-list] [@bridgenode/mcp on AI Agents Listing] [GitHub Stars]

Built for AI agents — no API keys, no registration, pay-as-you-go with Solana USDC via x402. Start free: free models and 2 free calls on paid models per client, no wallet required.

BridgeNode MCP wrapper — stdio MCP server that proxies requests to the remote https://bridgenode.cc/mcp (streamable-HTTP) with automatic x402 payment (Solana USDC). No API keys, no registration.

Installation (one-command)

bash
claude mcp add bridgenode -s user -- npx -y @bridgenode/mcp@latest

Or manually (Claude Code, Cursor, other MCP clients):

json
{  "mcpServers": {    "bridgenode": {      "command": "npx",      "args": ["-y", "@bridgenode/mcp@latest"]    }  }}

Configuration (.env)

bash
# Solana wallet private key (base58) — agent's payment wallet# Optional — Solana wallet private key (base58). Needed only for PAID calls:# free models and the 2 free trials work without it.BRIDGENODE_WALLET_KEY=
# Spending policy (fail-closed)BRIDGENODE_MAX_PER_CALL=0.05BRIDGENODE_DAILY_CAP=1.0

Tools

  • chat_completions — AI inference (free models run without payment; the first 2 calls to a PAID model are free per client, then x402 payment is automatic). Accepts OpenAI-style tools / tool_choice — the model's answer then carries tool_calls (the wrapper passes arguments through unchanged)
  • list_models — model list (free)
  • get_price_estimate — estimate cost of a chat request (free)

How it works

  1. Local MCP client launches the wrapper via stdio
  2. Wrapper connects to https://bridgenode.cc/mcp (streamable HTTP)
  3. On tools/call receiving 402 (payment required) — the wrapper automatically signs the payment with BRIDGENODE_WALLET_KEY and retries the request
  4. Spending policy checks every payment BEFORE signing (fail-closed)

Links

Python packages

Prefer Python? The same BridgeNode toolkit is on PyPI:

Community

⭐ Star this repo — it helps AI agents discover BridgeNode.

来源:README.md,提交 a0cfa36

工具

3

chat_completions

Chat completions with x402 payment

Send a chat completion request to any supported model. Free models are served without payment; the first calls to a PAID model are also free per client (free trials), after which x402 payment (Solana USDC) is required — an unpaid call returns 402 with the exact price; retry with _meta["x402/payment"]. Responses are non-streaming. Pass either mode (auto/eco/premium) or model (explicit id) — one of the two is required; model wins if both are sent. Use this tool to generate text; inspect models and prices first with the free list_models tool. Start with a FREE model (no payment at all) and keep max_tokens >= 200 — a thinking model spends part of that budget on reasoning, so a smaller limit can return an empty answer.

输入参数modemodeltoolsmessages必填max_tokenstool_choice

list_models

List available models and prices

List available models and their prices — free, no payment, no authentication required. Read-only: no state changes; data is served from the server's local config, so repeated calls return identical results (idempotent). Accepts no parameters: the input schema is an empty object, and any arguments passed are ignored. Calling it without arguments returns the complete catalog with per-token prices; there is no filtering, pagination, or configuration. Use this tool to inspect models and prices before calling the paid chat_completions tool. Same data as GET /v1/models (§5.2). Do not use it to generate text (use chat_completions) or to estimate a specific request's cost (use get_price_estimate).

无需输入参数。

get_price_estimate

Estimate the cost of a chat request

Estimate the USDC cost of a chat completion request before paying — free, no payment, no authentication required. Read-only: no state changes and no external calls; the estimate is computed locally from server pricing config, so repeated calls with identical inputs return identical results (idempotent). Use this tool to check the exact price for a given model/mode, messages, and max_tokens before calling the paid chat_completions tool. Provide either mode (auto/eco/premium routing) or model (explicit id, mutually exclusive with mode); one of the two is required — if both are sent, model wins. mode values: auto = cheapest model fitting the context, eco = cheapest available, premium = best model.

输入参数modemodelmessages必填max_tokens

版本历史

3
  1. v1.8.1最新Oct 11, 2026
  2. v1.7.0Oct 5, 2026
  3. v1.6.0Sep 16, 2026

评分

已满足适用检查项的 100%

该评分衡量此收录条目的完整程度,依据注册中心条目、端点与公开来源计算得出。它不是对服务器本身的评价,也不表示服务器安全或当前可用:可连通性是带日期的检查结果,评分也不包含安装量或用户评价。

  • 近 30 天内可连通必需25 分
  • 已声明认证方式必需10 分
  • 已列出工具必需15 分
  • 描述必需15 分
  • 图标10 分
  • 文档15 分
  • 许可证10 分