Meta is accelerating its bet on enterprise artificial intelligence with the imminent launch of a new version of its Muse Spark model, codenamed Watermelon. According to Alexandr Wang, the company's AI director, this update will bring substantial improvements in programming capabilities and agent systems, aiming to directly compete with market leaders like OpenAI and Anthropic.

What is Watermelon and why is it important?

Watermelon is not just a minor update. According to internal sources cited by Business Insider, this model has already reached the level of GPT-5.5, OpenAI's flagship model. To achieve this, Meta has allocated significantly more computational capacity than its predecessor, reflecting the intense competition to improve performance and expand capabilities in complex tasks such as code generation and automation through intelligent agents.

The announcement comes at a key moment for Meta, which seeks to close the gap with other players in the generative AI market and strengthen its presence in the enterprise segment. In a post on X, Wang acknowledged that the upcoming Muse Spark update will include major improvements in programming and agents to be more competitive against other leading models.

Impact on the enterprise market

For organizations, these improvements could have significant implications. According to Pareekh Jain, principal analyst at Pareekh Consulting, a more competitive model from Meta would intensify competition, reduce AI adoption costs, and offer an additional alternative to OpenAI and Anthropic. "If offered as an open-weight or low-cost model, it could make AI-based programming assistants more affordable while improving data control and reducing dependence on a single vendor," Jain says.

This context is especially relevant at a time when code review is being transformed by AI, and where GPU shortages and high licensing costs are making access to the most advanced capabilities expensive. A cheaper alternative driven by Meta could be attractive to companies looking to expand AI use without skyrocketing their technology budgets.

Beyond foundational models

The launch of Watermelon coincides with strategic moves by Meta that point to a more ambitious vision. Among them, attempts to acquire Manus and other initiatives that aim to turn Meta into a platform for building native AI applications and agents. Charlie Dai, principal analyst at Forrester, notes that "Meta wants to go beyond foundational models and become a platform for building native AI applications and agents."

Initiatives like Pocket, although consumer-oriented, show Meta's interest in lowering barriers to creating native AI software. However, the biggest opportunity lies in enterprise adoption: enabling business users to create workflow automations, agents, and lightweight applications with less technical expertise. This aligns with Meta's strategy to strengthen its presence in enterprise AI, including new business plans tied to cloud infrastructure.

Risks and challenges

Despite expectations, enterprise adoption is not guaranteed. Meta will have to demonstrate that its technological advances translate into consistent results in real corporate environments. "Meta must demonstrate superior code generation quality in real-world scenarios, reliable agent execution, strong security and governance capabilities, and a dynamic developer ecosystem," warns Dai.

Additionally, the international context adds new variables. Outside North America, geopolitical and regulatory considerations are influencing model selection, creating opportunities for alternatives. Meta needs to deliver compelling results, establish local partnerships, and maintain continuous innovation that connects with developers and businesses. In this regard, AWS's lessons on agentic AI with OpenTelemetry can be useful for understanding integration challenges.

What to expect from the launch?

Wang has announced that the new model will soon be available through Meta AI and via a new API, a move that will reinforce the company's strategy to attract both developers and enterprise customers. The question now is not only whether Watermelon will manage to compete with the most advanced models, but whether Meta can turn that technical improvement into a sufficiently solid business proposition to challenge OpenAI, Anthropic, and other providers currently leading the corporate AI race.

For IT professionals, this represents an opportunity to explore new tools that could optimize contract and commission management or improve security in virtualized environments. The evolution of enterprise AI is redefining the technology landscape, and Meta aims to be a key player in this transformation.

Original source: ComputerWorld. Analysis and adaptation by ForgeNEX.