Meta Delays Launch of 'Muse Spark' API, Impacting AI Monetization

A large logo installed in front of Meta Platforms' headquarters in Menlo Park, California. [Photo=AP]

Meta has reportedly postponed the launch of its latest artificial intelligence (AI) model's API multiple times. As the company continues to invest heavily in AI infrastructure, the delays are increasing pressure on its monetization efforts.

According to the Wall Street Journal on June 3, Meta has repeatedly delayed the release of the API for its latest AI model, 'Muse Spark.' Alexandre Wang, Meta's Chief AI Officer, had indicated to developers in April that the launch was imminent, but no confirmed release date has been established.

The API serves as a gateway for developers to integrate Meta's AI model into their applications or services. Companies like OpenAI and Anthropic have generated revenue by selling access to their APIs. For Meta, launching the developer API is crucial to recouping its investments in its AI model.

Meta initially planned to release the API in April alongside the unveiling of Muse Spark. However, the schedule was pushed to May due to identified bugs during testing and the need for additional infrastructure. The release date has since been further delayed to June.

In response to inquiries from the Wall Street Journal, Meta stated, "We are testing the API with partners and plan to launch it within this month." A company spokesperson added, "We know that people want the API, and we look forward to providing it."

This delay is intertwined with Meta's financial burden from its AI investments. The company plans to invest up to $145 billion in AI infrastructure this year. CEO Mark Zuckerberg has outlined a goal to create AI agents for both individuals and businesses.

Meta is also seeking ways to generate revenue from its investments. The company has announced plans to test a Meta AI subscription service and is considering a cloud computing business utilizing its surplus AI infrastructure capacity. However, the delayed API release means that a key monetization pathway remains unopened.

Muse Spark is the model that powers Meta's AI chatbot and related features. Previous AI models released by Meta were available for developers to download and use as open-source. In contrast, Muse Spark is the first model for which the design and software files have not been made publicly available.

Internal evaluations at Meta indicate that Muse Spark performs comparably to models from OpenAI and Anthropic. However, with the API release delayed, general developers still have limited access to utilize the model in their services.

* This article has been translated by AI.

AJP