Analyst(s): Nick Patience
Publication Date: May 19, 2025
The enterprise AI market has witnessed unprecedented acceleration in open model development throughout 2024 and early 2025, fundamentally challenging assumptions that proprietary models would maintain indefinite performance advantages. Meta’s Llama 3 release marked a watershed moment, while major cloud providers positioned themselves as neutral platforms offering both proprietary and open models.
Key Points:
- Meta, Mistral AI, and China’s DeepSeek have accelerated open-source model development, challenging the dominance of proprietary foundation models from OpenAI and Anthropic.
- Major cloud providers are increasingly positioning themselves as neutral platforms, offering access to both proprietary and open models, suggesting a multi-model future rather than a winner-take-all.
- Enterprise adoption patterns reveal a pragmatic split: open models for flexibility and customization, proprietary models for managed services, and cutting-edge capabilities.
Overview:
The enterprise AI market has witnessed unprecedented acceleration in open model development throughout 2024 and early 2025, fundamentally challenging the assumptions that proprietary models would maintain indefinite performance advantages. Meta’s release of Llama 3 models in April 2024 marked a watershed moment, demonstrating that open-source models could achieve near-parity with leading proprietary solutions. This was followed by significant contributions from Mistral AI, IBM, and NVIDIA, while China’s DeepSeek and Alibaba introduced competitive open-weight models that further intensified global competition.
The Open Source Revolution Gains Momentum: Meta’s Llama 3 release in April 2024 represented more than just another model launch – it fundamentally challenged the assumption that proprietary models would maintain indefinite performance advantages. The 8B and 70B parameter variants quickly established new benchmarks for open model performance, demonstrating capabilities that rival many proprietary alternatives. OpenAI’s surprising announcement in April 2025 of plans for an open-source model release signals recognition of the shifting market dynamics.
Global Competitive Landscape: The rise of international open-weight models has added geopolitical dimensions to the open versus proprietary debate. China’s DeepSeek and Alibaba have released models that not only compete on technical merit but also serve as demonstrations of national AI capabilities. The availability of high-quality open models from diverse geographical origins has democratized access to advanced AI capabilities and reduced dependence on a small set of Western proprietary providers.
Cloud Providers as Platform Orchestrators: Perhaps the most significant strategic development is the evolution of major cloud providers into AI platform orchestrators. AWS, Google Cloud, and Microsoft Azure have transcended their traditional infrastructure roles to become curators of AI model ecosystems. By offering both proprietary and open models through unified platforms, they’ve positioned themselves as the neutral ground where the open versus proprietary battle will ultimately be decided.
Enterprise Adoption Paradox: Despite the technical capabilities of open models, enterprise adoption reveals a more nuanced picture. When ChatGPT was launched in November 2022 and alternatives emerged in 2023 from cloud providers, most enterprises jumped at the chance to at least experiment with them. While enterprises often lean toward proprietary models, developers frequently prefer open-source models for their flexibility and control. Open-source models allow developers to inspect the code, fine-tune parameters, and deploy without ongoing licensing fees.
Figure 1: The Importance of Open Source to Developers
Economic Models: The economics of AI deployment favor different approaches for different use cases. Open models excel in scenarios requiring customization, transparency, or cost optimization for high-volume inference. Proprietary models remain attractive for cutting-edge capabilities, managed deployment, and scenarios where vendor support is critical.
Conclusion: The evolving AI landscape suggests that the open versus proprietary debate will not result in a single winner but rather in a more nuanced ecosystem where both approaches serve distinct needs and preferences. The future likely belongs to organizations that can skillfully navigate this dual landscape, leveraging open source models for flexibility, customer choice, and innovation while utilizing proprietary models for managed services and cutting-edge capabilities.
The full report, Open Source vs. Proprietary AI: Revolution or Just Another Market Split?, is available via subscription to the AI Software & Tools Practice IQ service from Futurum Intelligence—click here for inquiry and access.
Futurum clients can read more about it in the AI Software & Tools Intelligence Platform, and non-clients can learn more here: AI Software & Tools Practice.
About the Futurum AI Software & Tools Practice
The Futurum AI Software & Tools Practice provides actionable, objective insights for market leaders and their teams so they can respond to emerging opportunities and innovate. Public access to our coverage can be seen here. Follow news and updates from the Futurum Practice on LinkedIn and X. Visit the Futurum Newsroom for more information and insights.
Author Information
Nick is VP and Practice Lead for AI at The Futurum Group. Nick is a thought leader on the development, deployment and adoption of AI - an area he has been researching for 25 years. Prior to Futurum, Nick was a Managing Analyst with S&P Global Market Intelligence, with responsibility for 451 Research’s coverage of Data, AI, Analytics, Information Security and Risk. Nick became part of S&P Global through its 2019 acquisition of 451 Research, a pioneering analyst firm Nick co-founded in 1999. He is a sought-after speaker and advisor, known for his expertise in the drivers of AI adoption, industry use cases, and the infrastructure behind its development and deployment. Nick also spent three years as a product marketing lead at Recommind (now part of OpenText), a machine learning-driven eDiscovery software company. Nick is based in London.
