$145 billion Spending Cap: Meta Repeatedly Delays Muse Spark API, Urgency to Monetize Through Closed Source

According to Sentinel Beating monitoring, Meta (META) has repeatedly postponed the launch of its latest AI model Muse Spark's API, with no confirmed release date to date. A spokesperson stated on Wednesday that they are currently testing the API with partners and plan to launch it in June 2026. The delayed release has raised questions from the public about Meta's ability to monetize its substantial AI investment. Meta's capital expenditure plan for this year amounts to $145 billion, and back in April 2026, a stock price plunge of over 5% was attributed to higher-than-expected spending.

As Meta's first proprietary model, Muse Spark refuses to disclose its design blueprint and software weights, allowing developers only to access it through the API. In internal benchmark tests, Muse Spark's performance has been on par with OpenAI and Anthropic's flagship models and has surpassed xAI's Grok in most tests. However, apart from a few third-party evaluation agencies, the majority of developers have yet to conduct tests within the Agentic Harness intelligent framework.

Unable to sell the API interface directly, Meta has resorted to introducing paid subscriptions on Instagram, WhatsApp, Facebook, and Meta AI bots to offset the costs. CEO Mark Zuckerberg revealed that they do not rule out the possibility of renting out excess computing power in the future through cloud services. Prior to Muse Spark, Meta had scrapped the release of the Behemoth model due to performance issues, following which they restructured the AI division and established the Superintelligence Lab (MSL) led by Alexandr Wang. Muse Spark is the result of research and development in the lab's classified section, TBD Lab.