mlco2 open-source ecosystem

Measure the environmental footprint of computing & AI

Two tools, one mission.

Code Carbon is a French non-profit building free, open-source tools for developers and researchers. CodeCarbon for local compute, EcoLogits for GenAI APIs — use both when your stack mixes them.

CodeCarbon

Local compute emissions

Track CO₂ from code running on hardware you control — your laptop, workstation, or cloud VM. Measures CPU, GPU, and RAM power, then applies regional grid carbon intensity.

pip install codecarbon

EcoLogits

GenAI API impacts

Estimate energy and environmental impacts when you call third-party GenAI APIs — OpenAI, Anthropic, Mistral, Google, and more. Attaches impact data to provider SDK responses.

pip install ecologits

CodeCarbon docs

EcoLogits docs

How It Works

1

Download package

2

Embed the code

3

Run and track

4

Visualize results

Get started in minutes

Install a package, paste a snippet, run. Both tools side by side — pick Python or CLI/API within each column.

CodeCarbon

Local compute emissions

Docs

Install

terminal

bash

pip install codecarbon

train.py

python

from codecarbon import EmissionsTracker

tracker = EmissionsTracker()
tracker.start()

# Your code here

emissions = tracker.stop()
print(f"Emissions: {emissions} kg CO₂")

Wrap your script with EmissionsTracker — start, run, stop.

Full docs →

terminal

bash

# Track without changing your code
codecarbon monitor --no-api -- python train.py

# Detect your hardware
codecarbon detect

Track any command without editing source files.

Full docs →

EcoLogits

GenAI API impacts

Docs

Install

terminal

bash

pip install ecologits

chat.py

python

from ecologits import EcoLogits
from openai import OpenAI

EcoLogits.init(providers=["openai"])
client = OpenAI(api_key="<OPENAI_API_KEY>")

response = client.chat.completions.create(
    model="gpt-4o-mini",
    messages=[{"role": "user", "content": "Tell me a funny joke!"}],
)

print(f"Energy: {response.impacts.energy.value.mean} kWh")
print(f"GHG: {response.impacts.gwp.value.mean} kgCO2eq")

Init once, then read impacts from the SDK response.

Full docs →

terminal

bash

curl -X POST "https://api.ecologits.ai/v1/estimations" \
  -H "Content-Type: application/json" \
  -d '{
    "provider": "openai",
    "model": "gpt-4o-mini",
    "output_token_count": 150,
    "request_latency": 1.2
  }'

Estimate impacts over HTTP — no Python install required.

Full docs →

What we are

Lightweight, easy-to-use

Python libraries

Open source, free &

community driven

Effective

visual outputs

Related tools & integrations

First-party projects from mlco2, plus community-built extensions on our open APIs. Each tool belongs to CodeCarbon or EcoLogits — pick the stack you use.