AI cloud provider Nebius has unveiled a platform designed to challenge the dominance of major technology companies in the artificial intelligence infrastructure space. The company announced Token Factory on Nov. 5, positioning the service as a comprehensive solution for businesses seeking to deploy open-source AI models at scale while maintaining enterprise-level reliability and control.
The platform supports all major open models currently available in the market, including DeepSeek, OpenAI’s GPT-OSS, Meta’s Llama, Nvidia‘s Nemotron and Qwen. Token Factory launched with immediate availability and compatibility for over 60 open-source models, while also providing customers the flexibility to host their own proprietary models on the infrastructure.
Current Nebius AI users will receive automatic upgrades to the new Token Factory platform, ensuring seamless transitions for existing clients utilizing the company’s services.
Direct challenge to industry leaders
The new offering places Nebius in direct competition with Amazon Web Services, Microsoft Azure and Google Cloud Platform, three of the most established players in cloud computing infrastructure. The company also faces challenges from emerging startups like Fireworks and Baseten, which provide similar services focused on AI model deployment and optimization.
Nebius emphasizes that Token Factory has been optimized specifically for efficiency, promising sub-second latency, autoscaling throughput and 99.9% uptime reliability. These performance guarantees extend even to demanding workloads that exceed hundreds of millions of requests per minute, according to the company’s technical specifications.
The platform provides enterprises with the computing power necessary to run leading AI models while maintaining the flexibility to switch between different models as business needs evolve. This approach addresses growing demand from organizations seeking to avoid vendor lock-in situations common with proprietary AI platforms.
Company emerges from major restructuring
Nebius split from Russian search engine Yandex in 2024 following international sanctions, establishing itself as an independent entity focused on AI cloud services. The company has emerged as one of the leading neocloud providers, operating data centers strategically located across the United States, Europe and Israel.
The infrastructure provider sells AI cloud capacity from these facilities, positioning itself to serve clients across multiple geographic regions while maintaining compliance with various regulatory requirements. This global footprint allows the company to offer services to enterprises operating in different markets with varying data sovereignty needs.
AI infrastructure providers like Nebius that combine software services with cloud offerings potentially enjoy wider profit margins compared to companies focusing solely on hardware infrastructure. However, company leadership has indicated that margin expansion remains secondary to customer acquisition and product diversification strategies.
Strategic vision beyond basic infrastructure
CEO and Chief Business Officer Roman Chernin emphasized during an interview that the company views Token Factory as essential to its growth strategy rather than simply a margin-boosting initiative. The executive explained that merely providing infrastructure proves insufficient for building a substantial enterprise in the current competitive landscape.
Chernin articulated the company’s ambition to become a large enterprise while avoiding classification as merely a utility provider. The leadership team recognizes that clients increasingly seek flexibility in their AI model deployments, moving away from closed ecosystem approaches toward more diversified model portfolios.
The platform addresses this trend by offering what the company describes as a scalable and reliable foundation that enables clients to seamlessly transition from their initial model choices to whatever solutions they require at scale. This flexibility becomes particularly valuable as AI technology continues evolving rapidly, with new models and capabilities emerging regularly.
Market dynamics drive platform development
The introduction of Token Factory reflects broader shifts occurring within the AI infrastructure market as businesses grapple with decisions about model selection and deployment strategies. Organizations increasingly recognize the limitations of committing exclusively to proprietary platforms controlled by individual vendors, driving demand for more open and flexible alternatives.
Nebius positions Token Factory as a solution addressing these concerns by supporting both established open-source models and custom implementations. This approach allows businesses to experiment with different models, compare performance characteristics and optimize their AI deployments based on specific use case requirements rather than platform limitations.
The platform’s emphasis on reliability and uptime addresses critical concerns for enterprise customers deploying AI systems in production environments where downtime or performance degradation can result in significant business impact. The 99.9% uptime guarantee positions the service competitively against established cloud providers that have built reputations on infrastructure reliability.
As the AI infrastructure market continues maturing, competition among providers intensifies around factors including performance, reliability, flexibility and cost efficiency. Token Factory represents Nebius’s attempt to differentiate itself across these dimensions while carving out market share from established players.
Source: Live Mint