Spanish Startup Multiverse Computing Secures $215M to Revolutionize AI with CompactifAI

Multiverse Computing, a Spanish tech startup, is making significant strides in the field of artificial intelligence with its groundbreaking CompactifAI technology. Recently, the company announced its successful $215 million Series B funding round, highlighting investor confidence in its potential to revolutionize AI cost structures. At the heart of this success is CompactifAI, a technology designed to compress large language models (LLMs) by up to 95% without compromising performance. By achieving such efficient compression, Multiverse Computing enables faster and more cost‑effective deployment of AI models on various platforms, including AWS and on‑premise systems. This advancement is particularly beneficial for smaller organizations and individual users who have been deterred by the high operational costs of traditional AI models [¹].

CompactifAI represents a leap forward in AI technology, leveraging tensor networks which mimic the advanced processes of quantum computing but operate on classical hardware. This innovative approach allows for the compression of complex data structures within deep learning models, significantly reducing their size while maintaining their performance integrity. The technology’s application to open‑source LLMs such as Llama and Mistral has broadened its accessibility, enabling users to experience powerful AI capabilities on standard devices like personal computers and smartphones. The ability to run these models on everyday hardware not only democratizes access to AI but also highlights the ingenuity behind Multiverse's vision of transforming AI economics [¹].

The implications of CompactifAI extend beyond mere cost savings, paving the way for a more inclusive AI landscape. By significantly lowering the costs associated with AI deployment, Multiverse Computing empowers a diverse range of industries to integrate AI into their operations. The potential savings—estimated to be between 50% and 80% on inference costs—could make AI a viable option for educational institutions, small businesses, and non‑profits that previously could not afford it. Furthermore, this technology supports the trend towards edge AI deployments, where models are run efficiently on local devices rather than centralized servers, thereby reducing latency and improving privacy [¹].

As Multiverse Computing continues to expand its reach, the company’s co‑founders, Román Orús and Enrique Lizaso Olmos, are at the helm, steering its strategic implementation of CompactifAI. Their combined expertise in tensor networks and financial systems underlies the company’s innovative edge. The startup's impressive patent portfolio and solid customer base, including industry giants like Iberdrola and Bosch, signal its robust position within the AI landscape. As future enhancements are introduced, such as compatibility with additional models and proprietary solutions, the impact of CompactifAI is poised to grow even further [¹].

Multiverse Computing has recently made headlines by securing an impressive $215 million in Series B funding to catapult its groundbreaking AI technology to new heights. Known for their innovative CompactifAI technology, the Spanish startup has demonstrated a remarkable ability to compress large language models (LLMs) by up to 95% without compromising on performance. This technological advancement is particularly exciting because it offers a pathway to dramatically reduce AI model costs, an area that has often posed a financial barrier to broader adoption across various industries. For more details, you can visit the.¹

The recent financial backing Multiverse Computing received underscores the growing recognition of the CompactifAI's potential in the tech world. The company plans to leverage this funding to enhance their offerings and expand their market reach. Currently, CompactifAI enables efficient operation of open‑source LLMs like Llama and Mistral on widely accessible platforms such as AWS or through on‑premise licenses. This capability allows models to function effectively even on everyday devices like personal computers and smartphones, thereby democratizing AI technology access. Learn more in the.¹

With 160 patents to its name, Multiverse Computing is not only securing financial backing but also making significant strides in intellectual property. Their technology, aimed at reducing the size and operational costs of AI models, is not just theoretical; it is already being applied by leading global companies such as Iberdrola, Bosch, and the Bank of Canada. The ability to integrate this advanced compression technology into existing systems indicates a demand for AI solutions that balance cost with performance. For a comprehensive overview, feel free to check the.¹

This strategic investment is a testament to the transformative potential of Multiverse Computing in the competitive field of AI model management. By focusing on open‑source LLMs and intending to extend their support to models like DeepSeek R1, the company is positioning itself as a pivotal player in the movement toward more accessible AI. The drive to make AI a standard resource, available as easily as electricity, could be realized sooner than expected, thanks in part to innovations like CompactifAI. To explore the implications of this funding and technology, visit the.¹

CompactifAI technology stands at the forefront of artificial intelligence innovation, enabling significant advancements in model efficiency and accessibility. Multiverse Computing, the creative force behind CompactifAI, has developed a groundbreaking approach to compress large language models (LLMs) by up to 95% without compromising performance. This achievement not only reduces operational costs but also democratizes AI by making powerful models accessible for use on standard hardware, from personal computers to smartphones. As highlighted in their recent funding success, where the company raised $215 million in Series B funding, Multiverse is well‑positioned to scale its technology and expand its impact [1](https://techcrunch.com/2025/06/12/multiverse‑computing‑raises‑215m‑for‑tech‑that‑could‑radically‑slim‑ai‑costs/).

The essence of CompactifAI lies in its innovative use of tensor networks, which simulate quantum computations on classical hardware. This technology excels in reducing the complexity and size of data structures within deep learning models, achieving compression without performance degradation [1](https://techcrunch.com/2025/06/12/multiverse‑computing‑raises‑215m‑for‑tech‑that‑could‑radically‑slim‑ai‑costs/). By focusing on open‑source LLMs, such as Llama and Mistral, CompactifAI provides efficient and cost‑effective alternatives for AI deployment, allowing companies to scale AI applications more sustainably.

One of the compelling features of CompactifAI is its potential to drastically reduce AI inference costs, offering savings between 50% and 80% [1](https://techcrunch.com/2025/06/12/multiverse‑computing‑raises‑215m‑for‑tech‑that‑could‑radically‑slim‑ai‑costs/). This financial advantage is crucial for enterprises looking to leverage AI capabilities without incurring prohibitive costs. For instance, the cost of running a compressed version of Llama 4 Scout can be as low as $0.10 per million tokens on AWS, a significant reduction compared to the uncompressed model [1](https://techcrunch.com/2025/06/12/multiverse‑computing‑raises‑215m‑for‑tech‑that‑could‑radically‑slim‑ai‑costs/).

The versatility of CompactifAI extends to its deployment options, making it an attractive solution for diverse applications and environments. It supports deployment on platforms ranging from cloud services like AWS to on‑premise installations, and it is efficient enough to run on devices as varied as personal computers and Raspberry Pis [1](https://techcrunch.com/2025/06/12/multiverse‑computing‑raises‑215m‑for‑tech‑that‑could‑radically‑slim‑ai‑costs/). This flexibility ensures that businesses can integrate advanced AI functions into their operations with minimal hardware investment.

As the technology landscape evolves, the implications of CompactifAI's advancements are significant, not only for AI's economic and operational framework but also for the broader societal adoption of AI. By making AI tools more accessible and reducing infrastructure dependency, CompactifAI has the potential to bridge the digital divide and enhance data privacy through on‑premise deployments [1](https://techcrunch.com/2025/06/12/multiverse‑computing‑raises‑215m‑for‑tech‑that‑could‑radically‑slim‑ai‑costs/). Its ongoing development is key to maintaining an equitable digital society where AI benefits are shared widely across all sectors.

The integration and support of large language models (LLMs) with CompactifAI technology present a transformative opportunity for optimizing AI functionalities across various platforms. Multiverse Computing's innovative approach compresses LLMs significantly without compromising their performance, a breakthrough achieved through tensor networks—a design inspired by quantum computing principles. Tensor networks allow the intricate overlays of data compression to mimic quantum computer operations on classical hardware. With CompactifAI, open‑source models—the likes of Llama and Mistral—are now operationally feasible on smaller local devices, including PCs and even mobile handsets, heralding a new shift in AI deployment.

CompactifAI's compatibility with open‑source LLMs such as Llama 4 Scout and Mistral Small 3.1 reflects the startup's strategic focus on expanding AI access by offering more efficient alternatives to proprietary models like those from OpenAI. This technology enables substantial reductions in operational costs, with claims of slashed inference expenses by 50% to 80%. For enterprises, this translates into tangible financial savings and operational efficiencies, particularly when leveraging AWS cloud services to deploy these compressed models. The groundwork laid by Multiverse Computing supports a broader trend where companies prioritize cutting‑edge, adaptive technologies to enhance AI cost‑effectiveness and performance.

Despite the advantageous cost structure, CompactifAI currently renders its services around open‑source models, avoiding proprietary technologies. Such constraints, while presently a limitation in bypassing existing giants like OpenAI, can be seen as a tactical approach in building a strong foothold in the burgeoning market of accessible AI solutions. By concentrating on open‑source models, Multiverse Computing not only supports technological democratization but also fosters an environment ripe for innovation and adaptation. This strategy allows developers to exploit AI capabilities within budget constraints, essentially rendering AI an egalitarian resource.

The future of LLM compatibility hinges on CompactifAI's ability to expand its support for more models and enhance its technology's applicability across different contexts. The promise held by this technology in reducing infrastructure costs and promoting the sustainability of AI initiatives can potentially revolutionize the AI landscape. Moreover, as organizations adopt these compressed models, there is an implicit push towards reinforcing data security and privacy, especially in environments opting for on‑premise solutions. Multiverse Computing's stride towards these goals underscores the growing importance of versatile AI solutions that meet diverse usage requirements while adhering to cost‑effectiveness and performance standards.

In the rapidly evolving realm of artificial intelligence, the cost efficiency of compressed models stands as a significant breakthrough, promising a paradigm shift in AI deployment and operation costs. Companies like Multiverse Computing are pioneering this movement with their innovative CompactifAI technology, which achieves up to an astonishing 95% reduction in the size of large language models (LLMs). This compression does not come at the expense of performance, thereby maintaining the model's efficacy while drastically cutting costs. As reported by TechCrunch, these compressed models can run efficiently on less powerful hardware such as PCs and mobile devices, which were previously unsuitable for such tasks due to their limited processing power.

The economic implications of such advancements in model compression are profound. According to Multiverse Computing, employing their compressed models could lead to a reduction in inference costs by 50% to 80%, allowing businesses to operate AI more affordably and at greater scale. This cost reduction opens the door for smaller companies and startups, previously hindered by the high expense of AI technology, to leverage AI solutions effectively. In the competitive landscape of AI service providers, this advantage could translate into increased adoption and innovation across various industries, as highlighted in their recent.¹

Furthermore, the ability to deploy these compressed models beyond cloud infrastructure and onto personal and edge devices is likely to spur significant advancements in edge computing. Companies are increasingly focusing on solutions that allow AI to run on‑site, thus enhancing privacy and reducing latency associated with cloud‑based operations. As seen with Multiverse Computing's capability to support deployment on devices such as Raspberry Pis and even mobile phones, the flexibility and reach of artificially intelligent systems are expanding significantly, offering new potential use cases for AI technologies.

The strategic focus on compressing open‑source LLMs primarily stems from their accessibility and the economic advantages they offer in contrast to proprietary models. Multiverse Computing's expertise has allowed them to optimize models such as Llama and Mistral which are accessible open‑source LLM options. This open‑source focus not only democratizes access to AI technology but also fuels collaborative innovation within the community, driving AI development at an unprecedented pace. By offering solutions that significantly cut costs, companies like Multiverse Computing are pivotal in leveling the playing field in the AI domain.

The evolution of technology is seeing significant strides with Multiverse Computing's CompactifAI technology offering a promising shift in the deployment of large language models (LLMs) on‑premise while ensuring compatibility with various devices. As the trend towards edge computing continues, the ability to host sophisticated AI models on‑premise allows organizations not just to save on cloud‑related costs, but also to enhance data privacy and control. CompactifAI technology proves versatile, efficiently compressing LLMs to some of the smallest footprints available, enabling deployment on conventional hardware like personal computers, mobile devices, and even compact hardware such as Raspberry Pi. This stands in stark contrast to conventional AI setups that typically require substantial cloud infrastructure, thus democratizing AI access and offering significant cost advantages.

Multiverse Computing's CompactifAI technology ushers in a new era of device compatibility, enabling the execution of AI models across a wide range of devices. Traditionally, implementing complex AI models required expensive, high‑performance hardware. However, with CompactifAI's model compression, even standard PCs and phones are now capable of running these models without compromising performance. This paves the way for enhanced AI functionalities to be integrated into everyday devices, fostering an environment where AI is more accessible and can be tailored to meet individual user needs. By optimizing the hardware demands, Multiverse empowers businesses of all sizes to leverage AI technologies without the burden of substantial infrastructure costs, thus expanding AI's applicability across industries.

The strategic focus on enabling on‑premise deployment of AI models reflects a broader move toward ensuring data sovereignty and reducing dependency on cloud services. Multiverse Computing's technology is particularly critical for sectors that prioritize data privacy and require control over their data ecosystems. The speed and efficiency of CompactifAI also mean that organizations can handle sensitive data in‑house, thus aligning with regulatory requirements surrounding data sovereignty. Moreover, on‑premise deployment offers resilience against potential cloud outages, providing a more robust AI infrastructure that enhances operational continuity.

Compatibility with a variety of hardware configurations signifies a transformative change in how AI technology can be integrated into existing systems. Business leaders can now deploy, manage, and manipulate AI models within their existing IT frameworks, reducing the need for costly hardware upgrades or cloud subscriptions. This capability not only bolsters the operational efficiency of organizations but also offers flexibility in application, allowing AI models to be instantly updated and improved as needs evolve.

Multiverse Computing was co‑founded by the duo of CTO Román Orús and CEO Enrique Lizaso Olmos, both of whom bring unique expertise to the table. Román Orús is renowned for his pioneering work with tensor networks, which are crucial for the company's groundbreaking CompactifAI technology. This innovative technique allows the compression of large language models up to 95% while maintaining performance, a significant advancement in AI technology (¹). Meanwhile, Enrique Lizaso Olmos, with his extensive background in mathematics and banking, complements this technical expertise with a strong strategic vision, steering the company through its significant Series B funding and expanding its market presence.

Together, Orús and Lizaso Olmos have successfully positioned Multiverse Computing as a leader in AI model compression technology. Their combined experience and visionary leadership have not only secured substantial venture capital investments, such as the recent $215 million funding round, but also established strategic partnerships with major corporations including Bosch and the Bank of Canada. The company's trailblazing approaches and patent holdings have attracted attention and respect across the technology and business sectors, setting the stage for further innovation and growth.

Under their leadership, Multiverse Computing's CompactifAI is making waves in the AI industry, not only for its technical merit but also for the strategic implications of its deployment. The ability to run lean, efficient AI models on everyday devices like PCs and smartphones has the potential to disrupt traditional computing paradigms, a vision championed by Orús and Lizaso Olmos in their relentless pursuit of innovation. Their commitment to the democratization of AI technology underscores the foundational mission of Multiverse Computing, resonating well with investors and customers alike.

The technology landscape is witnessing a significant shift towards more efficient AI and machine learning models, largely driven by companies like Multiverse Computing. With their innovative CompactifAI technology, Multiverse Computing has successfully demonstrated the ability to compress large language models by up to 95% without compromising on their performance. This advancement holds the potential to drastically reduce AI operational costs, making sophisticated AI capabilities more accessible to a broader range of industries [¹].

Investment in technologies that improve AI efficiency has surged, reflecting the growing importance of cost‑effective AI solutions in today's market. Multiverse's $215 million funding round is a testament to the strong investor confidence in AI innovations that promise to cut costs and enhance performance. Such investments are pivotal for companies aiming to lead in the AI‑driven economy by providing scalable and affordable solutions [source].

The emphasis on developing alternative compression techniques for large language models is becoming more pronounced. As companies strive to optimize model architectures and explore new methods like quantization and pruning, there is a significant focus on supporting open‑source models. This focus not only helps reduce costs but also supports the growing trend toward making AI tools more available and flexible for various applications [source].

An increasing number of companies are looking into deploying smaller, more efficient AI models on edge devices. This movement toward edge AI reflects the need for operations that conserve resources while still maintaining powerful computational abilities. Multiverse's CompactifAI technology supports this trend by enabling these models to operate on lower‑power devices, broadening their applicability in diverse environments [source].

CompactifAI, a groundbreaking technology developed by Multiverse Computing, has stirred significant interest among experts in the fields of AI and technology. According to Roman, co‑founder and managing partner at Bullhound Capital, CompactifAI is poised to revolutionize how local large language models (LLMs) are utilized in devices ranging from cars to laptops and satellites. He emphasizes that the integration of this technology by hyperscalers could drastically reduce costs and enhance efficiency, paving the way for increased compute usage and facilitating private and sovereign cloud deployments, which could fundamentally alter the landscape of AI implementation [1](https://news.crunchbase.com/venture/quantum‑multiverse‑computing‑startup‑raises‑funding/).

Tuan Tran, who serves as the President of Technology and Innovation at HP, believes that Multiverse's innovative approach offers a pathway to improved AI applications by enhancing performance and ensuring personalization, while also maintaining privacy and cost efficiency for businesses of all sizes. Tran's insights underscore the multifaceted benefits of AI model compression, which not only reduces operational costs but also aligns with trends in privacy and customization [5](https://siliconangle.com/2025/06/12/multiverse‑computing‑bags‑215m‑quantum‑inspired‑ai‑model‑compression‑tech/).

While many experts are optimistic, Holger Mueller from Constellation Research Inc. calls for caution, pointing out the necessity of validating CompactifAI's effectiveness outside the realm of open‑source models. His perspective highlights the ongoing need for independent testing and verification to ensure these compressed models meet diverse industry needs and succeed in broader applications, thus stressing the importance of empirical evidence in validating technological claims [5](https://siliconangle.com/2025/06/12/multiverse‑computing‑bags‑215m‑quantum‑inspired‑ai‑model‑compression‑tech/).

Damien Henault, Managing Director at Forgepoint Capital International, is notably enthusiastic about CompactifAI, terming it a 'quantum leap' in AI deployment. He argues that this technological advance could lead to smarter, cheaper, and more environmentally friendly AI solutions, significantly influencing how AI is deployed across various sectors. Henault's endorsement reflects a broader industry trend towards sustainable AI development that balances performance with ecological considerations [8](https://thequantuminsider.com/2025/06/12/multiverse‑computing‑raises‑215‑million‑to‑scale‑technology‑that‑compresses‑llms‑by‑up‑to‑95/).

The introduction of Multiverse Computing's CompactifAI technology has been met with an intriguing mix of enthusiasm and skepticism. The promise of reducing AI operational costs by compressing large language models up to 95% while maintaining performance has sparked significant interest. This innovation, outlined in their recent $215 million Series B funding, is seen as a revolutionary step towards democratizing AI on a global scale. Indeed, the ability to run sophisticated AI models on standard hardware such as PCs and smartphones is perceived as a major leap forward, providing unprecedented access to AI capabilities [TechCrunch](https://techcrunch.com/2025/06/12/multiverse‑computing‑raises‑215m‑for‑tech‑that‑could‑radically‑slim‑ai‑costs/).

However, this excitement is tempered by a degree of caution among the public and industry observers. There are calls for external validation of the claimed compression and cost‑efficiency metrics before wider adoption can occur. Given that CompactifAI currently supports only open‑source models, debates about market reach and the potential exclusion of proprietary models like OpenAI's tools continue to swirl. Moreover, there are concerns about the long‑term sustainability and overarching impact of this technology on market dynamics, especially in terms of scalability and ethical implications related to increased AI deployment [TechCrunch](https://techcrunch.com/2025/06/12/multiverse‑computing‑raises‑215m‑for‑tech‑that‑could‑radically‑slim‑ai‑costs/).

In parallel, the discourse around CompactifAI hasn't been devoid of environmental concerns. The prospect of vastly increased AI usage on consumer devices raises questions about energy consumption and ecological impacts. Additionally, technology experts have pointed out that while the move towards personalized AI applications is positive, it requires strict guidelines to prevent potential misuse, calling for careful navigation through ethical and privacy considerations. The enthusiasm for CompactifAI's disruptive potential is palpable, but its journey will require navigating these complex technical and ethical landscapes [VMblog](https://vmblog.com/archive/2025/06/12/multiverse‑computing‑raises‑215m‑to‑scale‑ground‑breaking‑technology‑that‑compresses‑llms‑by‑up‑to‑95.aspx).

The dawn of multiverse technology represents a paradigm shift in how we perceive and interact with artificial intelligence (AI). With the advent of technologies like CompactifAI, developed by Spain's Multiverse Computing, the economic implications are profound. By compressing large language models (LLMs) by up to 95% while maintaining their performance, CompactifAI promises to significantly slash inference costs by 50% to 80%, as indicated in a recent.¹ Such advancements democratize access to high‑performing AI systems, previously confined to well‑funded tech giants due to high deployment costs.

The unparalleled success of Multiverse Computing, as demonstrated by its recent $215 million funding, marks a significant stride in the evolution of artificial intelligence technology. CompactifAI, their pioneering innovation, promises to revolutionize how large language models are utilized by dramatically reducing costs and resource requirements while maintaining high‑performance levels. This breakthrough opens up a plethora of economic opportunities by enabling more industries to incorporate powerful AI solutions without the historically prohibitive expenses associated with large‑scale AI deployment. As such, CompactifAI not only democratizes access to AI technology but also foreshadows a more inclusive and competitive digital economy.

Furthermore, the societal benefits heralded by CompactifAI are profound. By making advanced AI models operational on everyday hardware like PCs and mobile phones, the technology promises to bridge the digital divide and enhance digital literacy across various demographics. However, this increased accessibility must be handled with caution to prevent misuse and ensure equitable distribution of benefits. Ethically guided development and deployment will be crucial for tapping into the full potential of CompactifAI without exacerbating existing digital inequalities.

Politically, CompactifAI's support for localized on‑premise AI model deployment represents a shift towards greater data sovereignty and privacy. Organizations can manage their data in ways that align with specific regulatory requirements, a factor that may discourage reliance on centralized cloud services. This runs parallel to growing global discussions around data privacy, sovereignty, and security. While CompactifAI has the potential to champion greater independence in the AI industry, careful regulatory oversight will be essential to prevent any adverse concentration of AI capabilities that could impact market dynamics.

Looking ahead, the future implications of Multiverse's advancements rest on continued innovation and scalability efforts. While the technology's current application is focused on open‑source models, expanding these capabilities to proprietary models can further amplify its impact. As more companies and industries adopt CompactifAI, the cumulative economic, social, and political benefits will begin to unfold more clearly. The commitment of Multiverse Computing to further refine and expand its groundbreaking technology will no doubt play a decisive role in shaping the future landscape of AI applications globally. For more on this transformative advancement, visit ¹ for detailed insights.