Meta’s layoffs-and-AI pitch meets its first real-world test: Muse Spark API still delayed

Meta Platforms is heading into a familiar CEO task, translating expensive AI spending into products developers can actually use, and doing it on a schedule that investors can track. A June 7 report focused on what it called the company’s first “real test” after earlier job reductions: the long-delayed developer API for Muse Spark, Meta’s newer model for its Meta AI assistant.

Earlier this year, Meta announced plans to cut about 10% of its workforce, or roughly 8,000 employees, with the layoffs set to begin May 20, according to reporting. The company also told workers it was seeking to run more efficiently while offsetting the cost of other investments. At the same time, Meta increased its spending outlook, raising its full-year 2026 capital expenditures to a range of $125 billion to $145 billion, citing higher component pricing and additional data center costs.

Muse Spark itself was positioned as the next step in Meta Superintelligence Labs’ effort to rebuild Meta’s AI stack. In an April 8 post, Meta said Muse Spark currently powers the Meta AI app and, and that it would gradually roll out across other Meta products. It also said it would open access to the underlying technology through a private preview via API to select partners, and that it hoped to open-source future versions.

But the developer-access piece has not landed cleanly. TheStreet, citing a Reuters report and people familiar with the matter, said Meta delayed plans for the Muse Spark developer API multiple times and, as of June 3, had no planned launch date. The holdup was attributed to “software bugs and infrastructure problems” found during testing. TheStreet also noted that Meta Chief AI Officer Alexandr Wang told developers on X two days after Muse Spark’s April debut that the API would be “coming soon,” without it appearing when expected.

Meta’s own stance, as described in the June 7 report, is that the company is testing the API with early partners and expects to release it “this month.” That response addresses what Meta is doing, but not why the timeline has slipped repeatedly from its earlier expectations. For investors, the issue is the timing: Meta’s layoffs were framed less as a general cost-cutting exercise and more as a resource reallocation tied to AI ambitions, while Muse Spark was presented as a flagship outcome of that effort.

From a business perspective, a developer API is more than a technical feature. It is the channel through which outside companies and builders can integrate a model into their own apps, products, and workflows. When such access is delayed, developers cannot reliably build and test integrations, and any monetization path that depends on ecosystem adoption becomes harder to forecast on a tight timetable, especially when spending guidance is large and explicit.

Still, key details remain unclear. Meta has not publicly laid out granular information about what specific bugs or infrastructure issues were found, whether performance concerns contributed to the delay, or how far along the API release process is internally. It is also not disclosed what revenue or usage targets, if any, Meta planned to hit immediately after launch. What to watch next is straightforward: whether Meta actually ships the Muse Spark API during June, whether early partners receive broader access without additional slips, and whether later updates connect the API rollout to measurable progress in how Meta’s AI investment is being converted into product value.