2.0× less energy than Muse Spark.
Muse Spark Environmental Impact
First model from Meta Superintelligence Labs — powers Meta AI
The figures below are for one quick question. A longer chat with follow-ups uses about 1.4 Wh, and an hour or two of agent work about 40.0 Wh.
One query, in context
Energy, carbon and water for a single use. Quick single question (one exchange)
About 1.3× a Google search and 48% of a minute of laptop use.
About 1.3× a Google search and 7% of 10 m by car.
About 15× a drop of water and 15% of a teaspoon of water.
- Where it runsMeta data centres (US)
- Grid carbon intensity384 g CO₂/kWh · 43% renewable
- Compared with a Google searchabout the same as a search (0.3 Wh)
- Water~1,351 queries to fill 1 litre
Where Muse Spark sits
#98 of 199 by energy per query. Lighter than 55% of all models. Lit dots are the same kind of model; click any dot to open it.
Want something lighter?
1.7× less energy than Muse Spark.
5.7× less energy than Muse Spark.
What a month of use adds up to
Pick how much you use it. Based on one quick use at a time, five days a week.
- Energy
- 43.3 Wh
- Carbon
- 16.2 g
- Water
- 80 mL
A month of 5 chats a day, 5 days a week. The energy is about 3.6× a phone charge and 39% of a kettle boil.
Work out your own →- Energy
- 173.2 Wh
- Carbon
- 65.0 g
- Water
- 320 mL
A month of 20 chats a day, 5 days a week. The energy is about 1.6× a kettle boil and 2% of a UK home's day.
Work out your own →- Energy
- 866.0 Wh
- Carbon
- 324.8 g
- Water
- 1.60 L
A month of 100 chats a day, 5 days a week. The energy is about 7.9× a kettle boil and 12% of a UK home's day.
Work out your own →How does Muse Spark compare?
Next to similar models. Bars use a log scale, so each step is a big jump.
Detailed breakdown
Energy
We estimate Muse Spark at ~0.4 Wh for a quick single question. Meta has published no energy data for the Muse family and no independent estimate exists, so this is a class-based figure: Muse Spark is a closed frontier-class model, and CLEER (Sustainable AI Group, Sept 2026) puts comparable 2026 flagships at ~0.4-0.5 Wh on current hardware. Muse Spark is embedded in Meta AI across WhatsApp, Instagram and Facebook, so much of its use is incidental — updated versions (1.1 to 1.3) shipped between July and September 2026.
Power source & carbon
Runs in Meta's own data centres; we apply the US average grid (~384 gCO2/kWh). Meta matches its electricity use with renewable purchases on an annual basis, but its location-based emissions are growing with its AI build-out, including new gas-backed campuses.
Water use
At ~0.74 mL per quick question, using a ~1.85 L/kWh combined water factor. Meta does not disclose per-model water use.
About Muse Spark
Muse Spark is a text and chat model from Meta, released in April 8, 2026. First model from Meta Superintelligence Labs — powers Meta AI. Each query uses 0.40 Wh of energy and produces 0.15 g of CO₂. That's roughly comparable to a traditional Google search.
These figures are estimates derived from hardware specifications and API benchmarks — Meta has not published official energy data for Muse Spark. Actual consumption may vary significantly depending on batching, quantisation, and infrastructure optimisations that we cannot observe from outside.
Muse Spark in context
What does your Muse Spark usage cost the planet?
Use the calculator to estimate your personal footprint based on how often you use Muse Spark.
Frequently asked questions
How much energy does Muse Spark use per query?
Each Muse Spark query consumes approximately 0.40 Wh of energy. This is about the same as a traditional Google search (~0.3 Wh).
What is Muse Spark's carbon footprint?
Based on the carbon intensity of Meta data centres (US), each query produces approximately 0.15 g of CO2. The grid in this region has a carbon intensity of 384 g CO2/kWh with 43% renewable energy.
How much water does Muse Spark use?
Each query consumes approximately 0.74 mL of water, primarily used for cooling the data centers that process the request.
How does Muse Spark compare to a Google search?
A Muse Spark query uses about the same as a Google search in terms of energy. A Google search uses approximately 0.3 Wh, while Muse Spark uses 0.40 Wh.
Technical details
- Architecture
- Natively multimodal Transformer with visual chain-of-thought
- Release date
- 2026-04-08
- Open source
- No