proxyllm-mcp
Click on "Deploy Server".
Wait a few minutes for the server to deploy. Once ready, it will show a "Started" state.
In the chat, type
@followed by the MCP server name and your instructions, e.g., "@proxyllm-mcp how much would I save on $500 monthly OpenAI spend?"
That's it! The server will respond to your query, and you can continue using it as needed.
Here is a step-by-step guide with screenshots.
proxyllm-mcp
MCP server for ProxyLLM, the OpenAI-compatible LLM gateway. It gives any MCP client (Claude Code, Claude Desktop, Cursor, Cline, Windsurf, OpenClaw) live model catalogs, plan-savings math, routing-key introspection and provisioning, and an account signup flow an agent can complete on its own: email OTP in, sk_ account token out, HTTP 402 with a checkout link until a human activates the membership.
Install
Claude Code:
claude mcp add proxyllm -- npx -y proxyllm-mcpCursor, Cline, Claude Desktop, or any client that takes a JSON server entry:
{
"mcpServers": {
"proxyllm": {
"command": "npx",
"args": ["-y", "proxyllm-mcp"]
}
}
}No configuration is required to start. Two optional environment variables save passing tokens per call:
PROXYLLM_ACCOUNT_TOKEN: ansk_(read+write) orrk_(read-only) account token for the management tools.PROXYLLM_ROUTING_KEY: apllm_routing key for key introspection and key-scoped model lists.
Related MCP server: Multi-Provider LLM MCP Server
Tools
Free, no account needed:
proxyllm_savings_calculator: monthly OpenAI API bill in, subscription tier + flat total + monthly savings out.proxyllm_list_models: OpenAI-shaped model list from the gateway.proxyllm_introspect_key: what a routing key can do (lanes, models, budget, whether a Codex subscription backs it).
Account flow (an agent can run this end to end):
proxyllm_signup: emails a 6-digit code to the operator's inbox. No captcha.proxyllm_verify_signup: exchanges the code for thesk_account token (returned once).proxyllm_account: account state; while unpaid it carries the checkout link and a message written to be relayed to the operator. Poll untilplanis"pro".
Provisioning (paid accounts):
proxyllm_create_routing_key: mints apllm_key, optionally wiring its provider chain (Codex subscription seats, self-hosted Claude Code bridges, metered API keys) in the same call.proxyllm_list_routing_keys: keys, plus the wireable lane ids withinclude_lanes.proxyllm_usage: 30-day usage summary.
Unpaid accounts get 402 payment_required on management tools by design; the body names the price ($129/mo flat), the checkout URL, and the 48-hour removal deadline for never-activated accounts. Error bodies are returned verbatim because they contain the next step.
Links
Agent auth manifest: https://proxyllm.ai/auth.md
OpenAPI: https://proxyllm.ai/openapi.json
Agent card: https://proxyllm.ai/.well-known/agent-card.json
MIT license.
Available Tools
9 toolsproxyllm_accountAccount state (poll until paid)A
GET /v1/organizations/me. Works while unpaid: plan is "free" plus a payment block until the operator activates the membership, then "pro". Poll this after handing over the checkout link.
Inputs: account_token.
proxyllm_create_routing_keyCreate a routing keyA
Mints a pllm_ routing key (returned once) for OPENAI_API_KEY-style use against the gateway. Optionally wires the provider chain in the same call; discover wireable lane ids with proxyllm_list_routing_keys include_lanes.
Inputs: mode, label, providers, account_token, +1 more.
proxyllm_introspect_keyIntrospect a routing keyA
GET /v1/key: the key's provider lanes, reachable models, whether a Codex subscription backs it, and monthly budget state. Authed by the routing key itself.
Inputs: routing_key.
proxyllm_list_modelsList modelsA
OpenAI-shaped model list from the gateway. No auth required; with a routing key it reflects that key's lanes, including the live Codex catalog.
Inputs: routing_key.
proxyllm_list_routing_keysList routing keys (and wireable lanes)A
Lists the account's routing keys. include_lanes adds the wireable lanes (connected Codex sessions, bridges, saved provider keys) with the ids provider chains reference.
Inputs: account_token, include_lanes.
proxyllm_savings_calculatorLLM spend: API vs subscription mathA
Given a monthly OpenAI API bill in USD, returns the ChatGPT/Codex subscription tier that absorbs that workload through ProxyLLM Codex Hosted, the flat total (subscription + $129 fee), and the monthly savings. Free, no auth.
Inputs: monthly_openai_bill_usd (required).
proxyllm_signupCreate a ProxyLLM account (email OTP)A
Starts autonomous account creation: a 6-digit code is emailed to the operator's inbox (no captcha, no browser). New accounts are unpaid and are removed after 48 hours unless a human activates the membership. Follow with proxyllm_verify_signup.
Inputs: email (required), source.
proxyllm_usage30-day usage summaryC
GET /v1/organizations/usage for the account.
Inputs: account_token.
proxyllm_verify_signupVerify the emailed code, receive the account tokenA
Exchanges the 6-digit code for an sk_ account token. The token is returned exactly once: store it (PROXYLLM_ACCOUNT_TOKEN). While the account is unpaid the response carries a checkout link to relay to the operator.
Inputs: code (required), email (required).
Tool Schema Changelog
Recent tool additions, removals, and schema changes observed during successful MCP inspections.
Loading tool schema changelog…
TDQS
Scored across 9 tools
Each tool targets a distinct resource or action: savings calculation, model listing, key introspection, signup steps, account details, key creation/listing, and usage. There is no overlap; even the sequential signup tools are separated by stage.
All tools share the 'proxyllm_' prefix and most follow a verb_noun pattern (list_models, create_routing_key, verify_signup). However, a few are noun-only (account, usage, signup) and savings_calculator is noun_noun, creating minor inconsistency.
The 9 tools are well-scoped for a gateway management service, covering signup, key lifecycle, model discovery, usage, and savings estimation. It is neither too thin nor overloaded.
The set covers the core account and key lifecycle comprehensively, including signup, verification, key creation/listing, introspection, usage, and models. Minor gaps exist such as no delete or update routing key, but they can be worked around.
Maintenance
Related MCP Connectors
Hosted MCP server for LLM cost estimation, model comparison, and budget-aware routing.
An MCP server that provides an API to LLMs to manage their JumpCloud resources.
MCP server providing access to the Scorecard API to evaluate and optimize LLM systems.
LLM cost optimization gateway, repo token-waste auditor & model arbitrage MCP server.
Related MCP Servers
- AlicenseNot gradedqualityDmaintenanceMCP server that provides tools to interact with the LiteLLM proxy API, enabling LLM completions, embeddings, image generation, and admin operations.MIT
- FlicenseAqualityDmaintenanceUnified MCP server for querying multiple LLM providers (Gemini, OpenAI, Anthropic, etc.) with advanced routing, cost optimization, and fallback resilience.2-
- AlicenseAqualityBmaintenanceMCP server for the LiteLLM proxy, enabling agents to administer resources like keys, teams, users, models, and proxy settings through risk-graded meta-tools.6MIT
- AlicenseBqualityBmaintenanceMCP server for OpenRouter's management API, enabling programmatic control of credits, inference keys, guardrails, and usage analytics from MCP-compatible clients.16123 npmMIT