• The Association presented this Monday at CEOE a document, annexed to its AI Gen White Paper, which analyzes in detail the role and advantages of SLM.
• These language models represent a strategic gateway to expand the use of AI to more companies, sectors and public services, facilitating the digital leap without cost or complexity barriers.

Madrid, December 12, 2025 – The Spanish Association for Digitalization, DigitalES, has published a new report dedicated to Small Language Models (SLM), one of the most promising areas within generative Artificial Intelligence (AI), as an annex to its Generative AI White Paper published last year.

The document, which will be presented next Monday at CEOE, analyzes in detail the role of SLMs, their advantages over large language models (LLMs) and their most relevant applications in key sectors. The objective of DigitalES, a business association that brings together the main companies in the technology sector in Spain, is to offer companies, administrations and society a clear framework to understand why this technology represents a unique opportunity to accelerate digitization in an efficient, secure and sustainable way.

WHAT ARE SLM AND WHY DO THEY MATTER?

Unlike massive language models such as GPT-4 or Gemini, SLMs have a much smaller number of parameters, allowing them to operate in resource-constrained environments, significantly reducing energy consumption and infrastructure costs.

This translates into key benefits:

• Efficiency and sustainability: lower carbon footprint and energy consumption.
• Accessibility: SMEs and organizations with fewer resources can deploy advanced AI.
• Privacy and security: possibility of local processing, avoiding sending sensitive data to the cloud.
• Specialization: high capacity to adjust to specific domains (finance, health, education, telecommunications).

A BOOMING MARKET

Interest in SLMs is not just theoretical. A report by the consulting firm MarketsandMarkets estimates that the global market for these models will reach $5.45 billion in 2032, up from $930 million in 2025, with a compound annual growth rate of more than 28%.

In Spain, the context is particularly favorable:

• According to official sources 21.1 % of companies with 10 or more employees reported using AI technologies in the last quarter of 2025 (INE), compared to 8 % of the EU average.
• Among SMEs, adoption grew by 36% in the last year, although the overall implementation rate remains low (2.9%).). (IndesIA – Barometer of AI adoption in Spanish SMEs 2025).
• Spain has the optimal conditions to become a technological hub capable of driving innovation in Europe. Forty-five percent of the country’s largest companies have already made significant investments in artificial intelligence (“Europe facing the challenge of AI: reinventing industries for a new one”, Accenture). An effort that is beginning to translate into tangible results: 70% of Spanish organizations report direct economic benefits from the use of these technologies(II European AI Barometer, EY).

SLMs therefore represent a strategic gateway to expand the use of AI to more companies, sectors and public services, facilitating the digital leap without cost or complexity barriers.

“SLMs represent a paradigm shift: they make it possible for companies, administrations and organizations of all sizes to leverage the potential of AI without the need for expensive infrastructures. This annex seeks to be a practical guide to understand their benefits, limitations and real applications” explains Beatriz Arias, digital transformation director at DigitalES.

AN UNDERSTANDABLE FRAMEWORK FOR UNDERSTANDING AND IMPLEMENTING SLM

Through specific industry examples, the DigitalES paper shows how SLMs are transforming the practice of artificial intelligence:

• Telecommunications: automatic call analysis, service personalization and real-time support chatbots.
• Banking and finance: fraud detection, transaction classification and customer assistants.
• Health: virtual triage systems, clinical documentation management and personalized reminders.
• Retail and e-commerce: post-sales chatbots, customer review analysis and intelligent product categorization.
• Corporate education and training: automatic tutorials, performance analysis and content customization.

These examples show that SLM is already a real and scalable tool for improving productivity, customer satisfaction and operational efficiency in multiple industries.

The report also addresses the technological challenges and opportunities posed by this trend: from model optimization and compression techniques, to the potential of SLMs in the development of edge computing, data protection and the consolidation of open source ecosystems. Finally, the paper invites reflection on the future of SLMs and their role in the democratization of artificial intelligence in Europe.

With this report, DigitalES provides companies, administrations and research centers with a practical and rigorous tool to understand and take advantage of the opportunities of Small Language Models. It is a document that not only informs, but also inspires to rethink how AI can be more inclusive, sustainable and adapted to the real needs of each sector.

The Annex on Small Language Models has been prepared by the Generative AI Working Group of DigitalES, led by Accenture, with the participation of experts from Deloitte, IBM, Nokia, NTT Data, Overlap and Quobis, as well as Accenture, as authors.

The report is available for download here .