Perplexity AI Environmental Impact

Energy Consumption

Perplexity AI consumes approximately 0.50 Wh per search query. This blended estimate accounts for: ~70% of queries routed to a small model (~0.1 Wh), ~30% using GPT-4o class (~0.43 Wh), plus ~0.3 Wh for web search augmentation (crawling, re-ranking, summarisation).

Power Source & Carbon

Perplexity uses a multi-cloud setup, primarily AWS. The web retrieval component adds additional compute beyond the LLM inference itself — each query involves network I/O, web crawling, and document processing.

Water Usage

At approximately 0.9 mL per search query, Perplexity has a modest water footprint. The blended nature of its queries (mostly small model routing) keeps consumption low.

How much energy does Perplexity AI use per query?

Each Perplexity AI query consumes approximately 0.50 Wh of energy. This is 2x more than a traditional Google search (~0.3 Wh).

What is Perplexity AI's carbon footprint?

Based on the carbon intensity of AWS US / Multi-cloud, each query produces approximately 0.18 g of CO2. The grid in this region has a carbon intensity of 450 g CO2/kWh with 25% renewable energy.

How much water does Perplexity AI use?

Each query consumes approximately 0.90 mL of water, primarily used for cooling the data centers that process the request.

How does Perplexity AI compare to a Google search?

A Perplexity AI query uses 2x more than a Google search in terms of energy. A Google search uses approximately 0.3 Wh, while Perplexity AI uses 0.50 Wh.

Architecture

Multi-model routing + RAG pipeline

Release date

2024-01-01

Open source

No