Cloud platform Salesforce on Monday announced that it launched the AI Energy Score with its partners, a benchmarking tool designed to evaluate and compare the energy efficiency of artificial intelligence models.

According to a statement, the company partnered with Hugging Face, Cohere, and Carnegie Mellon University to create this tool. It will be the first AI model developer to disclose the energy consumption of its proprietary models under this new framework.

The AI Energy Score, unveiled at the AI Action Summit, comes amid growing concerns about artificial intelligence’s environmental toll.

As AI systems become increasingly complex and energy-intensive, the initiative seeks to establish a universal standard akin to the Energy Star rating for household appliances, offering a clear and trusted benchmark for sustainability.

“Transparency like that offered by the AI Energy Score is crucial,” said Ariane Thomas, global tech director of sustainability at L’OREAL Group in a statement. “By openly sharing energy consumption data, companies can collectively implement ecodesign practices and minimize AI’s environmental footprint.”

A Standard for Sustainable AI

The AI Energy Score is designed to assess models across 10 common AI tasks, including text generation, image generation, and summarization. Its key features include:

  • Standardized Energy Ratings: A universal framework for measuring AI model energy efficiency.
  • Public Leaderboard: A ranking system showcasing efficiency scores for 166 AI models, including Salesforce’s SFR-Embedding, xLAM, and SF-TextBase.
  • Benchmarking Portal: A submission platform where AI developers can have their models evaluated.
  • Recognizable Energy Labels: A 1- to 5-star rating system, with five stars indicating the highest efficiency.

The project has already received recognition from the French Government and the Paris Peace Forum for its potential to drive sustainable AI practices. By publicizing energy efficiency data, the AI Energy Score encourages developers and enterprises to prioritize environmentally responsible models.

Salesforce’s Commitment to AI Sustainability

Salesforce has also highlighted its efforts to minimize AI’s energy impact, particularly through Agentforce, an autonomous AI platform integrated into its ecosystem.

Unlike traditional AI deployments that require energy-intensive model training for each user, Agentforce leverages optimized small language models and agentic reasoning to deliver high performance with lower computational costs.

“Reducing AI energy consumption lowers operational costs, optimizes infrastructure, and enhances long-term sustainability,” said Suzanne DiBianca, EVP and Chief Impact Officer at Salesforce. “We are proud to work with industry leaders to build a more transparent AI ecosystem.”

For instance, Salesforce’s SFR-RAG model, which specializes in extracting precise facts and citing sources, is designed for efficiency, requiring less computational power than traditional large-scale language models.

A Global Call for Responsible AI Development

With AI’s rapid expansion, industry leaders acknowledge the urgency of balancing technological advancements with sustainability. The AI Action Summit, which convenes representatives from over 100 countries and private sector and civil society leaders, serves as a crucial forum for shaping AI’s future.

Bruno Bonnell, general secretary for investment in France, emphasized the initiative’s global importance: “The AI Energy Score exemplifies the mission of the AI Action Summit by tackling a pressing societal issue—AI transparency and sustainability—through bold innovation and global collaboration.”

The AI Energy Score is expected to set a precedent for greater accountability in AI development, encouraging more companies to disclose their models’ environmental impact and adopt energy-efficient solutions.

As AI continues to evolve, initiatives like these signal a shift toward a more sustainable and responsible technological future.