Confier une tâche à un agent dans le navigateur

Name: Give a task to a browser agent
Creator: Projet Celsius
License: https://creativecommons.org/licenses/by/4.0/

Projet Celsius

Database entryUsesLevel 3 of 6 useUse, level 3 of 6Used to calculate 8 entries, level 4Calculated with 6 entries, levels 1 and 2Database structure

Give a task to a browser agent

An agent that browses websites on its own to complete a task, in 15 steps. Projet Celsius estimate.

"Analyse an image" accounts for 30% of emissions in this scenario.
100% of the calculation rests on database assumptions; it does not measure actual use.
Out of scope: server manufacturing, model training, the network and the user's device.

Compare

Calculated with

Used to calculate

Breakdown of emissions by item

5.3 gCO2e in total: 12 Wh of electricity × 458 gCO2e/kWh (servers worldwide).

Inputs of the calculation 5

PUE of the data centre, Analyse an image, Read 1,000 tokens, Re-read from cache, Write, with reasoning

Annual emissions from giving a task to a browser agent, by frequency

Your frequencyPeriod

Read the values at85 times a year

Reading: at 85 uses per year, "Browser agent" weighs as much as charging a phone once a week. Carbon footprint (CO2e) over one year, by number of uses. Each use is a straight line; "Browser agent" is highlighted, with the chosen parameters, and compared with 6 reference uses at their default settings; the solid line is one phone charge per week over a year, and the point where the task reaches it. Servers worldwide. Calculation from the Projet Celsius AI Impact Database.

How the calculation is built

At each step it reads a screenshot and re-reads its history. The number of steps follows published web agent trajectories, 13 on average.

Comparison with the other uses

The topic "Code and agents"

Generate a function0.54 gCO2e, 0.1 times the value of "Browser agent"
Have code reviewed0.59 gCO2e, 0.1 times the value of "Browser agent"
Browser agent5.3 gCO2e

↓ The heaviest use in the topic, "Coding agent hour": 310 gCO2e, 58 times more

Emissions of one use, linear scale, 3 of the 6 uses in the topic, those of the same order of magnitude (at most 25 times between the smallest and the largest value). For calculated values, servers worldwide.

Characteristics

Unit	1 task
Default location	world, 458 gCO2e/kWh (Ember 2025, life cycle); 11 other locations to choose from
Counted	the model reading and writing tokens, and data centre cooling
Not counted	server manufacturing, model training, the network and the user's device
Status	estimated

All the characteristics identifier, version, rule for the value, sources, licence

Identifier	IA-0220, projetcelsius.com/en/ai/database/IA-0220/
Version	0.1 of 2 November 2026
Category	Uses, a request to AI
Choice of the selected value	database calculation from the entries this one is made of, at the quantities of the unit
Sources	calculation of the database
Licence	CC BY 4.0

Emissions by server location

Emissions associated with electricity (CO2e)

Poland6.8 gCO2e
World5.3 gCO2e
United States4.5 gCO2e
European Union2.4 gCO2e
France0.48 gCO2e

× 14 between Poland and France, for the same electricity

Ember 2025, over the life cycle. The same use consumes 12 Wh in each location.

Assign a task to a browser agent · result per task · estimated · one server location per row.Included: the model reading and writing tokens, and data centre cooling. Out of scope: server manufacturing, model training, the network and the user's deviceFind this case: projetcelsius.com/en/ai/database/tasks/browser-agent-task/Projet Celsius calculation, AI Impact Database, version 0.4 of 13 November 2026 · IA-0220, entry in version 0.1 · engine 0.3.1 · CC BY 4.0.

Usage parameters

Advanced settings model class, reasoning, calibration

Model classReasoningCalibration

Result for the chosen case

Location of the servers

For a carbon footprint

The row for a carbon footprint

Item	purchases of services (GHG Protocol scope 3, category 1); regulatory BEGES, category 4
Quantity	1 task
Factor	5.3 gCO2e per task, IA-0220 version 0.1, according to the grid at the location (location-based), servers worldwide
Total	5.3 gCO2e · not counted: server manufacturing, model training, the network and the user's device

Breakdown of emissions by item

Emissions associated with electricity (CO2e)

Analyse an image1.6 gCO2e · 30 %
Re-read from cache1.4 gCO2e · 27 %
Write, with reasoning1.1 gCO2e · 20 %
Read 1,000 tokens0.72 gCO2e · 14 %
PUE of the data centre0.48 gCO2e · 9.1 %

Assign a task to a browser agent · result per task · estimated · servers worldwide.Included: the model reading and writing tokens, and data centre cooling. Out of scope: server manufacturing, model training, the network and the user's deviceFind this case: projetcelsius.com/en/ai/database/tasks/browser-agent-task/Projet Celsius calculation, AI Impact Database, version 0.4 of 13 November 2026 · IA-0220, entry in version 0.1 · engine 0.3.1 · CC BY 4.0.

Breakdown by item

Quantity, unit value, contribution and nature of the data for each item

Item	Quantity	Unit value	Contribution	Share	Nature of the data
Analyse an image	× 9.83 after calibration, 15 before	0.35 Wh	3.5 Wh	30 %	assumption · Zhang and Nam
Re-read from cache	× 98.3 after calibration, 150 before	0.032 Wh	3.1 Wh	27 %	assumption
Write, with reasoning (without reasoning)	× 2.46 after calibration, 3.75 before	0.96 Wh	2.4 Wh	20 %	assumption
Read 1,000 tokens	× 4.94 after calibration, 7.54 before	0.32 Wh	1.6 Wh	14 %	assumption
PUE of the data centre	× 1.10 on the rest		1.1 Wh	9 %	convention

Entries that use it (8)

Autonomous agent, 1 month, Office worker, 1 year, Executive assistant, 1 year, Manager or executive, 1 year, Consultant, 1 year, Salesperson, 1 year, ChatGPT Plus, 1 year, ChatGPT Pro, 1 year

Sources of the calculation

4 publications, for this entry and those used to calculate it, with the chosen parameters.

The result is calculated by Projet Celsius. The sources below support its uses, its factors and its conversion; they do not all measure this total.

Uses and assumptions

These publications shed light on AI tasks and uses. They do not measure the carbon footprint of the case calculated.

Anthropic · 2026
Vision, Claude API documentation ↗ (new tab)
Read 1,000 tokens
Role and limits : Anthropic
What the publication covers. tokens of an image according to its size: one patch of 28 × 28 pixels per token, 1,296 for a megapixel
What it does not cover. electricity
Reading by the database. The publisher's billing rule
- Analyse an image → : Read 1,000 tokens →
  The photo cut into about 1,100 image tokens, read as text
  Scope of the figure used. server, excluding the data centre
  Limit. The quantity and the share entrusted to AI are database assumptions, not a measured footprint.
- Analyse an image → : Read 1,000 tokens →
  Length convention: a photo sent to an AI
  Scope of the figure used. A one-megapixel photo, cut into image patches: 765 tokens at OpenAI for GPT-4o, 1,032 at Google, 1,229 for recent OpenAI models, 1,296 at Anthropic
  Limit. This publication supports the quantity of text or content processed, not its carbon footprint.
Zhang and Nam · June 2026
HANSEL: Extracting Breadcrumbs from Web Agent Trajectories for Interactive Verification ↗ (new tab)
Analyse an image
Role and limits : Zhang and Nam
Author. Yujin Zhang and Daye Nam
What the publication covers. 45 web agent trajectories: 13.16 steps and 6.02 pages visited per task on average
What it does not cover. electricity
Reading by the database. A small sample, which gives the order of magnitude of an agent's steps
- Agent in the browser → : Analyse an image →
  A screenshot read at each step, 15 steps: published web agents take 13 on average
  Scope of the figure used. image reading on the server; outside the data centre
  Limit. The quantity and the share entrusted to AI are database assumptions, not a measured footprint.

Energy and factors

The references of the selected values and of the entries used in this calculation.

Watershed · Aug. 2026
Estimating GHG Emissions from AI Use: Framework for Corporate-Level Measurement ↗ (new tab)
Read 1,000 tokens · Write 1,000 tokens · Data centre PUE
Role and limits : Watershed
Author. Bistline, Ulissi, Davis, Glidden, Joyce, Li, Mohsenin and Suh (Watershed)
What the publication covers. factors per 1,000 tokens read and written at the server before PUE (host and standby machines included), PUE separate (1.10 for a large operator, 1.56 for an unknown operator)
What it does not cover. clean measurements, the model and the hardware
Reading by the database. Default factors proposed for review (public data as of June 2026); for a large operator, taken from Google's measurement published in 2025 (Elsworth et al.), which already contains the data centre overhead; constant per token, they underestimate long contexts and agents according to the authors
- Read 1,000 tokens →
  Reference for the variant "Medium model, Watershed default factor"
  Scope of the figure used. Default factor for 1,000 tokens read, medium model, large operator, at the server before PUE (Watershed table 2, public data as of June 2026)
  Limit. A default value that Watershed derives from Google's measurement (0.24 Wh for the median text query of the Gemini app), with about 500 tokens per query and a written token costing 3 times a read token; Watershed does not give the split between read and written tokens. The 0.24 Wh already include the overhead of Google's data centre (0.02 Wh, 8%): with 375 tokens read and 125 written, the simplest reading, the factor contains it, and the PUE of 1.10 applied afterwards counts it a second time (about 9%). Values proposed for review, final at the next version of the framework
- Write 1,000 tokens →
  Reference for the variant "Medium model, Watershed default factor"
  Scope of the figure used. Default factor for 1,000 tokens written, medium model, large operator, at the server before PUE (Watershed table 2, public data as of June 2026)
  Limit. 3 times the reading factor, based on the ratio between writing and reading selected by Watershed (Patel et al., Splitwise, 2024). Same derivation and same caveat as for reading: Google's 0.24 Wh already include about 8% of data centre. Values proposed for review, final at the next version of the framework
- Agent in the browser → : Data centre PUE →
  The data centre, convention C3
  Scope of the figure used. building cooling and losses
  Limit. This figure covers part of the calculation; it does not measure the whole entry.
- Data centre PUE →
  Reference of the selected value
  Scope of the figure used. Default factor for a large operator, from the server to the data centre; 1.56 for an unknown operator
  Limit. A default value, not a measurement

Conversion into carbon

The source of the electricity factor of the selected location.

Worldwide: 458 gCO2e/kWh, life cycle.

Ember · 2026
Global Electricity Review 2026, and Our World in Data based on Ember ↗ (new tab)
Grid electricity
Role and limits : Ember
Author. Ember, Our World in Data
What the publication covers. electricity generation over the life cycle, by country
What it does not cover. guarantees of origin
Reading by the database. The same source for all countries
- Grid electricity →
  Conversion of electricity into carbon: worldwide
  Scope of the figure used. the production of electricity over the life cycle, including power plant construction and fuels
  Limit. An average for the chosen country, not the electricity contract or the data centre of a specific service.

Nature of the data

Published readings, conventions, assumptions

What this figure rests on

Projet Celsius assumptions 100 %

A share counts as an assumption as soon as a quantity set by Projet Celsius (a frequency, a share done with AI, a headcount) enters its calculation, even if the rest rests on published measurements.

Calculation assumptions

Quantities, nature and source of each assumption

Analyse an image : × 9.83
A screenshot read at each step, 15 steps: published web agents take 13 on average
assumption · Zhang and Nam · Analyse an image
Read 1,000 tokens : × 4.94
The instruction and the new text of each page, 500 tokens per step
assumption · Read 1,000 tokens
Re-read from cache : × 98.3
The agent's instructions and its history, 10,000 tokens re-read from cache at each step
assumption · Re-read from cache
Write, with reasoning : × 2.46
The action chosen at each step and its justification, 250 tokens
assumption · Write, with reasoning

Methodology conventions

With their alternatives

The full methodology

The data centre	Server electricity is multiplied by 1.10 for the data centre · alternatives: × 1.72 from chips to the full service at Google, never to be combined
The cache	A token re-read from the cache costs 10% of a token read · alternatives: bounds 1% and 25%
The default location	Without other information, the servers are in the world, at the world average electricity · alternatives: the United States, where most of the large assistants run
Electricity, location-based, on a life cycle basis	The kilowatt-hour is that of the servers' country, on a life cycle basis, from a single source for all countries · alternatives: market-based, with purchase contracts, shown alongside when published
Two significant figures	A calculated value is rounded to two significant figures, a published value is reproduced with the precision of its source; the calculation keeps full precision
The formula	the sum of what makes up the entry, then CO2e: electricity (Wh) multiplied by the carbon intensity of the server location (gCO2e/kWh), divided by 1,000
Calculate a case without an entry	The formula, the default values and a step-by-step example

History

Entry in version 0.1 of 2 November 2026, in the database in version 0.4

The changelog

2 November 2026Version 0.1 of the entry, as of the date of its current value.
31 October 2026Creation of the database, in its first version (0.1).

Citation

Projet Celsius, AI Impact Database, version 0.4 of 13 November 2026, "Assign a task to a browser agent", IA-0220 (entry in version 0.1 of 2 November 2026), projetcelsius.com/en/ai/database/IA-0220/ · CC BY 4.0.

Download

JSON Markdown