Payperwork Inference MCP

An MCP server providing OpenAI-compatible inference for frontier models like GPT-6.1, Claude, and Gemini. It enables agents to draft, summarize, extract, and classify data using a pay-per-call model settled via USDC on Base, with no API keys or accounts required.

Visit Payperwork

Endpoint and access

No sign-in to connect and list tools. Individual tools may still need a vendor account, credentials, or permissions; check the documented prerequisites.

Vendor evidenceStreamable HTTPNo sign-inEndpoint checked 2026-11-01Operated by Payperwork

You need:Any OpenAI client plus an x402 fetch wrapper ( @x402/fetch , the Python or Go x402 SDKs) is the whole integrationfor MCP, an x402-aware client such as the Cloudflare Agents SDK's withX402Client

MCP endpoint (Streamable HTTP)

https://payperwork.app/ai/mcp

Vendor documentation

What you can do with it

Drafting content via chat completions
Summarizing text from multiple models
Extracting structured data with JSON output
Classifying inputs using frontier models

Source and maintenance

Vendor evidenceRegistry 0.4.0

mcp-registry: app.payperwork/ai@0.4.0 Project website

Protocol and capabilities

Server: payperwork-ai 0.4.0
MCP protocol: 2025-06-18
Tools: 2 advertised
Resources: Not advertised
Resource templates: Not advertised
Prompts: Not advertised

Server-advertised capabilities

{
  "tools": {
    "listChanged": true
  }
}

Server-provided instructions

Untrusted server output shown for inspection. It is not Koskamo guidance and is excluded from knowledge search.

Chat completions on frontier models (Claude, GPT, Gemini, Grok), paid per call in USDC over x402. The chat tool quotes its price in the 402 for your exact request; models is free. Security model: every call is priced before it runs (the 402 states the amount) and charged only if it succeeds; the paying wallet is the identity; no account or key exists to revoke. Full statement: GET / (JSON) or /llms.txt. Inference: prompts go to the model's provider and are not stored by Payperwork; a failed model call is not charged; thinking models spend part of max_tokens on reasoning.

Tools

2 advertised

chat

Advertised by the server during the last successful anonymous check.

Advertised read-only

Chat completion on a frontier model: GPT-6.1 Sol and Terra, GPT-5.5, Claude Sonnet 5.5 and Haiku 5.5, Gemini 3.8 Flash, Grok 4.7, GPT-6 Luna (default anthropic/claude-haiku-5.5; see the models tool). Same shape as OpenAI chat: messages, optional max_tokens (default 1024, max 8192), temperature, tools. Price is quoted per call from your input size and max_tokens: (estimated input tokens x input rate + max_tokens x output rate) x 1.4 + $0.0005, minimum $0.002; the 402 states the exact amount. Charged only if the model call succeeds. Thinking models (GPT-5.5, Gemini, Grok, gpt-oss) spend part of max_tokens on reasoning: give them room, or set reasoning_effort (default low). Prompts go to the model's provider and are not stored by Payperwork. No streaming.

chat input schema

{
  "type": "object",
  "$schema": "https://json-schema.org/draft/2020-12/schema",
  "required": [
    "messages"
  ],
  "properties": {
    "stop": {
      "anyOf": [
        {
          "type": "string"
        },
        {
          "type": "array",
          "items": {
            "type": "string"
          }
        }
      ]
    },
    "model": {
      "type": "string"
    },
    "tools": {
      "type": "array",
      "items": {}
    },
    "top_p": {
      "type": "number",
      "maximum": 1,
      "minimum": 0
    },
    "messages": {
      "type": "array",
      "items": {
        "type": "object",
        "required": [
          "role",
          "content"
        ],
        "properties": {
          "name": {
            "type": "string"
          },
          "role": {
            "type": "string"
          },
          "content": {},
          "tool_calls": {},
          "tool_call_id": {
            "type": "string"
          }
        },
        "additionalProperties": {}
      },
      "maxItems": 200,
      "minItems": 1
    },
    "max_tokens": {
      "type": "integer",
      "maximum": 8192,
      "minimum": 1
    },
    "temperature": {
      "type": "number",
      "maximum": 2,
      "minimum": 0
    },
    "tool_choice": {},
    "response_format": {},
    "reasoning_effort": {
      "enum": [
        "none",
        "low",
        "medium",
        "high"
      ],
      "type": "string"
    }
  }
}

chat annotations

{
  "title": "Chat",
  "readOnlyHint": true
}

models — Models

Advertised by the server during the last successful anonymous check.

Advertised read-only

List the available models with provider rates, the pricing formula and limits. Free.

models input schema

{
  "type": "object",
  "$schema": "https://json-schema.org/draft/2020-12/schema",
  "properties": {}
}

models annotations

{
  "readOnlyHint": true
}

Prompts

Not advertised

No prompt definitions available.

Resources

Not advertised

No resource definitions available.

30-day availability

Aggregated from read-only remote checks; this is not a tool-execution or security guarantee.

Successful checks: 1 of 1 (100%)
Average latency: 636 ms
95th percentile latency: 636 ms

Maintenance history

Public check and source-refresh events. Editorial and ownership records remain private.

Nov 1, 2026, 12:11 PM UTC · probe · passed · 636 ms · metadata changed · 2 tool(s)

Connect from your client

Each snippet registers the same endpoint; pick the client you use.

This server is operated by the named third party. Listing and endpoint checks are not an endorsement or a security audit. Review the operator's permissions, terms, and data handling before connecting. How listings are checked and corrected.

Claude Code

Run this once in a terminal. It registers the remote server. Use --scope project to write .mcp.json for the repo.

Claude Code configuration

claude mcp add --transport http payperwork-inference-mcp https://payperwork.app/ai/mcp

Codex

Run this once. Codex stores MCP servers in ~/.codex/config.toml (or .codex/config.toml in a trusted project).

Codex configuration

codex mcp add payperwork-inference-mcp --url https://payperwork.app/ai/mcp

Cursor

Merge this server into .cursor/mcp.json in your project, then enable payperwork-inference-mcp in your MCP settings.

Cursor configuration

{
  "mcpServers": {
    "payperwork-inference-mcp": {
      "url": "https://payperwork.app/ai/mcp"
    }
  }
}

Cline

Merge this into cline_mcp_settings.json (Cline → MCP Servers → Configure).

Cline configuration

{
  "mcpServers": {
    "payperwork-inference-mcp": {
      "type": "streamableHttp",
      "url": "https://payperwork.app/ai/mcp",
      "disabled": false
    }
  }
}

VS Code

Merge this server into .vscode/mcp.json. Use MCP: List Servers to start payperwork-inference-mcp and inspect its tools.

VS Code configuration

{
  "servers": {
    "payperwork-inference-mcp": {
      "type": "http",
      "url": "https://payperwork.app/ai/mcp"
    }
  }
}

Gemini CLI

Run this once. It writes .gemini/settings.json in the current project; pass --scope user for ~/.gemini/settings.json.

Gemini CLI configuration

gemini mcp add --transport http payperwork-inference-mcp https://payperwork.app/ai/mcp

Claude Desktop

Merge this into claude_desktop_config.json (macOS: ~/Library/Application Support/Claude/, Windows: %APPDATA%\Claude\), then restart Claude Desktop.

Claude Desktop configuration

{
  "mcpServers": {
    "payperwork-inference-mcp": {
      "url": "https://payperwork.app/ai/mcp"
    }
  }
}

Windsurf

Merge this into ~/.codeium/windsurf/mcp_config.json, then refresh MCP in Cascade. Windsurf remote servers use serverUrl, not url.

Windsurf configuration

{
  "mcpServers": {
    "payperwork-inference-mcp": {
      "serverUrl": "https://payperwork.app/ai/mcp"
    }
  }
}

Let your agent find its tools

Find servers like Payperwork Inference MCP from your agent. Add Koskamo once to search the index for endpoints, access requirements, and client setup. Public discovery needs no key.

Connect Koskamo Explore workflows

See all 4,391

Do you run Payperwork Inference MCP?

Prove ownership with a DNS record, a well-known file, or your Official MCP Registry entry, and your edits publish immediately. Proofs are re-checked weekly; admins can still hide or remove a listing.

Claim this listing