Claude
- Open Add custom connector ↗
- Name it Carbon of Compute and paste the URL
- Click Add, then ask away
https://mcp.carbonofcompute.com/mcp
Carbon footprint of AI models
Carbon of Compute estimates the greenhouse gas and water footprint of running AI language models, per 1,000 output tokens or per task. Each figure is split into four parts: the electricity the hardware draws, the manufacturing of that hardware spread over its working life, the datacenter facility plant, and optionally a share of model training. Footprint results come with low and high values.
You can ask for one model's footprint, compare several models under the same assumptions, and change the electricity grid region, datacenter efficiency (PUE), hardware lifetime, utilization and cooling regime. The app also returns the manufacturing footprint of AI accelerators (NVIDIA H100, H200, B200, A100, AMD MI300X) split into logic die, HBM memory, package and board; an 8-GPU server line by line; and datacenter facility factors.
Intended for engineers, sustainability analysts, researchers and journalists who need a sourced estimate with its assumptions shown. Footprint and ranking answers link to the same view on carbonofcompute.com and to the method report. Hardware manufacturing figures come from REEL LCI, whose foreground inventory is built from public sources and linked to licensed ecoinvent 3.12 background results.
Limitations: figures are estimates built from public evidence for a reference deployment, not a provider's own measurement of its service. The low-high span is not complete uncertainty. Separately executed routing, safety and tool services are not consistently covered, and the datacenter building shell is excluded. The app only reads published data; it does not take actions or access any account.
What is the carbon footprint of Llama 3.3 70B per 1,000 output tokens?
https://mcp.carbonofcompute.com/mcphttps://mcp.carbonofcompute.com/mcpmcp.carbonofcompute.comThe AI reads these server instructions before using this connector, every time.
Carbon of Compute estimates the carbon (gCO2e) and water (mL) footprint of AI language models per unit of output: operational electricity, hardware manufacturing, datacenter facility plant and optional training. finds a model id, covers one model, compares models under the same assumptions, and the manufacturing tools return accelerator, server and facility factors. Footprint and ranking answers state their scenario and low and high values, and link to the same view on carbonofcompute.com.
None in the store listings or on the server
Carbon footprint of AI models. Carbon of Compute estimates the greenhouse gas and water footprint of running AI language models, per 1,000 output tokens or per task. Each figure is split into four parts: the electricity the hardware draws, the manufacturing of that hardware spread over its working life, the datacenter facility plant, and optionally a share of model training.
Open its listing in the app's connector directory and connect. No sign-in is needed. Step-by-step setup for each app is on this page.
The Carbon of Compute MCP server URL is https://mcp.carbonofcompute.com/mcp. Paste it as a custom connector in Claude, ChatGPT, Gemini or Grok, or add it to Claude Code, Cursor or any MCP client. Setup steps.
Yes. Carbon of Compute is published or run by REEL LCI (Scope Creep, LLC), so it is the official MCP server. See other official connectors.
It is listed in 1 store: ChatGPT. Apps where it is not listed can usually still add it as a custom connector.
It is ranked 14,679 of 15,293 in the ChatGPT directory. A rank is the store's own order, not a usage count. How we track ranks.
It has 8 tools: 8 read data. See every tool and its inputs.
No. All of its tools only read data.
None that we found. None in the store listings or on the server.
Here are examples from its store listing. Type one into your AI after you connect.
No. Its answers come back as text in the chat. Connectors with interactive UI.
This page updates every day. To ask about Carbon of Compute from Claude, ChatGPT or a coding agent, connect the Agent Ready MCP server: it tracks ranks, tools and new listings across the AI plugin stores.