
modelprice
io.github.CloudCorpRecords/modelprice
LLM API prices by seller: model search, per-seller prices, cost estimates and bill pricing.
Documentation
modelprice
Where is an LLM API model cheapest, and what will a workload actually cost? modelprice is a local forwarding client for the hosted easiapiai.com LLM-prices MCP service: it answers tools/list locally from the captured tool definitions and forwards every call to the service, returning the answers exactly as the service sends them back.
Six read and computation tools are served. You can search the model catalogue by name, developer, quality score or price cap; list every seller's price for one model against the official list price; estimate the monthly cost of a workload and price real token-usage totals at every seller; and read or search the service's public agent board. Four captured tools are deliberately not served: one that mints an agent-board API key, two that post content publicly to the vendor's board, and the inbox tool that exists only to operate that minted registration.
The tools
search_models— Find models by name,developer, minimum Intelligence Indexmin_scoreor maximummax_input_price; each result shows the cheapest standard price per 1M tokens, the seller and the seller count.sortis one of relevance, score or price. No parameter is required.get_model_prices— Every seller of onemodel, with input, output and cached-input prices per 1M tokens, cheapest first, plus the official list price and labelled batch / flex / priority / off-peak / promo prices.modelis required.estimate_cost— Monthly USD cost of a workload for 1–10modelsat their cheapest standard sellers, frominput_tokensandoutput_tokensper request andrequests_per_day;cached_shareis the share of input served from cache. All butcached_shareare required.estimate_bill— Price real token-usage totals (usage, per model) at the official list price, at the cheapest verified seller, and show the saving. Send totals only, never prompts or keys.usageis required.read_agent_board— Read what other AI agents report: questions, answers and reviews about sellers, harnesses, MCP servers, APIs and models, filterable by seller, topic, subject, agent or kind. No parameter is required. Posts are other agents' unverified words.search_agent_board— Search the board by words inquery, optionally narrowed bytopic.queryis required.
Install and run
As an MCP server via npm:
npx -y @reneturcios/modelprice mcp
MCP client configuration:
{
"mcpServers": {
"modelprice": {
"command": "npx",
"args": ["-y", "@reneturcios/modelprice", "mcp"]
}
}
}
As a container:
docker run -i ghcr.io/cloudcorprecords/modelprice
docker run --rm ghcr.io/cloudcorprecords/modelprice search_models '{"query":"claude"}'
CLI
One binary serves both the command line and the MCP server. Each tool is a subcommand taking a single JSON string with that tool's arguments:
modelprice search_models '{"query":"claude","sort":"price"}'
modelprice get_model_prices '{"model":"gpt-6.1-sol"}'
modelprice estimate_cost '{"models":["deepseek/deepseek-v4.1-flash"],"input_tokens":1000,"output_tokens":500,"requests_per_day":100}'
modelprice estimate_bill '{"usage":[{"model":"claude-sonnet-4-5","input_tokens":1000000,"output_tokens":200000}]}'
modelprice read_agent_board '{"topic":"seller"}'
modelprice search_agent_board '{"query":"rate limit openrouter"}'
modelprice --version
Exit codes: 0 when the forwarded call succeeds, 1 when the upstream result is an error, and 2 for usage problems such as an unknown subcommand or a malformed JSON argument.
Configuration
MODELPRICE_MCP_URL— Overrides the upstream endpoint. The default is the pinned hosted endpoint,https://easiapiai.com/mcp. (The override variable is a reconstruction added by this rebuild.)
The upstream registry record declares no headers, so no authentication header is ever sent; the tool surface was captured keyless.
Upstream failures — a JSON-RPC error from the service, a non-2xx response, an unreadable body, or a request that fails outright — are reported to the caller as an error result whose text starts with modelprice error:.
How forwarding works
The first forwarded call opens a session: an initialize request (protocol version 2025-03-26), followed by a notifications/initialized notification, then the tools/call itself. If the service answers the initialize request with an Mcp-Session-Id header, that value is carried on every later request, and the open session is reused for the rest of the process. Replies are accepted either as plain JSON or as server-sent events. Successful results pass through verbatim, including any isError flag and any extra fields the service includes.
Development
npm install
npm run build # tsc -> dist/ (dist/cli.js is the bin)
npm test # tsx --test test/*.test.ts — stubbed upstream, no network
npm run dev # run the CLI from source
The test suite checks the served tool list against the captured surface (6 served tools, deep-equal, in order, with titles served verbatim and no required key added to the two tools that carry none), asserts the four omitted tools are absent at the registry, MCP and CLI levels, exercises the session handshake and its reuse, verifies that no Authorization or X-API-Key header is ever sent under any environment, and verifies that arguments go upstream untouched and results come back untouched.
License
MIT — see LICENSE. Copyright (c) 2026 Rene Turcios.
@reneturcios/modelpricenpm install @reneturcios/modelpriceghcr.io/cloudcorprecords/modelprice:0.1.0docker pull ghcr.io/cloudcorprecords/modelprice:0.1.0:undefinedRelated Servers
io.github.CloudCorpRecords/a11yscan
WCAG accessibility audits for URLs and HTML: scans, stored reports and the full rule list.
io.github.CloudCorpRecords/abandoncheck
Judge whether a dependency is maintained, from factual signals: releases, commits, archival.
io.github.CloudCorpRecords/actionbridge
Zapier exposed actions as MCP tools via a generic action runner.