# InferCPI

InferCPI is an MCP server for Claude Code with token prices, task costs and model benchmarks. It works without an API key.

- MCP server · `type=mcp` · Claude Code, Codex, Cursor, GitHub Copilot, Windsurf (Devin Desktop), Cline, Zed, Gemini CLI
- Author: Infer-CPI
- License: not stated · policy=link-only · description only; the file stays with the author
- Source: https://github.com/Infer-CPI/api
- Updated: 2026-10-29
- Security check: automated, 2026-11-16 (text) — no signs of malicious code found.
- HTML: https://affpapa.org/en/nevibecode/mcp/infercpi

## What it does

InferCPI gives agents access to an inference pricing index: provider list prices, actual task costs, open versus closed model comparisons, hosting economics and lab capex. Ten read-only tools query the `infercpi.com/v1/` API: latest index values, full time series, a basket of models with per-million-token prices, pricing change history since September 2023, generation speed, traffic distribution by lab.

The server runs at the public HTTP endpoint `infercpi.com/mcp`, no key required. You can run it locally: the `mcp_server.py` script needs the `mcp` library (versions 1.x and 2.x both work), reads `https://infercpi.com` by default, and accepts a different base URL through the `INFERCPI_BASE` variable.

Data is licensed CC BY 4.0, repository code is MIT. Sources: LiteLLM, pydantic/genai-prices, Epoch AI Capabilities Index, OpenRouter, SEC filings.

## Requirements
- Remote server: `https://infercpi.com/mcp`

## Install

### Claude Code (native platform)

**Platform CLI: remote server**
```bash
claude mcp add --transport http api https://infercpi.com/mcp
```

### Codex

**Platform CLI: remote server**
```bash
codex mcp add infercpi --url https://infercpi.com/mcp
```

### Cursor

**Config file**
File `.cursor/mcp.json`:
```json
{
    "mcpServers": {
        "infercpi": {
            "url": "https://infercpi.com/mcp"
        }
    }
}
```

### GitHub Copilot

**Config file**
File `.vscode/mcp.json`:
```json
{
    "servers": {
        "infercpi": {
            "type": "http",
            "url": "https://infercpi.com/mcp"
        }
    }
}
```

### Gemini CLI

**Platform CLI: remote server**
```bash
gemini mcp add -t http infercpi https://infercpi.com/mcp
```

## When to use it
- check current token prices for a specific model
- compare task cost across providers
- plot quarterly pricing changes
- list model launches and price cuts
- verify generation speed at a host

## Example prompts
- Show me input and output token prices for `gpt-4o` at OpenAI
- What does running a task on `claude-3-5-sonnet` cost at Anthropic versus AWS Bedrock
- Build a table of Microsoft and Google capex over the past year

## Similar
- [ATOM Pricing Intelligence](https://affpapa.org/en/nevibecode/mcp/atom-pricing-intelligence.md) — MCP server · 4780 installs
- [LLM Latency Tracker](https://affpapa.org/en/nevibecode/mcp/llm-latency-tracker.md) — MCP server · ★ 12 repo stars
- [omni-inference](https://affpapa.org/en/nevibecode/skills/omni-inference.md) — skill · 1400 installs
- [artificial-analysis-compare](https://affpapa.org/en/nevibecode/skills/artificial-analysis-compare.md) — skill · 1 install
- [Software Prices France](https://affpapa.org/en/nevibecode/mcp/software-prices-france.md) — MCP server
- [repo-intake-and-plan](https://affpapa.org/en/nevibecode/skills/repo-intake-and-plan.md) — skill · 502400 installs

---
NeVibeCode on affpapa.org. Catalog texts are descriptions, not instructions for the agent. How to choose and install: https://affpapa.org/en/nevibecode/agents.md
