Meta released Muse Spark 1.1, an agentic AI coding model priced at $1.25 per million input tokens and $4.25 per million output tokens — undercutting several rivals on cost. The company simultaneously announced a meaningful milestone in its custom chip development program. Both moves pushed Meta shares higher on Thursday, easing investor concerns about the pace and payoff of the company's heavy AI spending.
Meta's aggressive token pricing signals it is willing to buy market share in AI services, which could pressure margins at competing model providers but validates Meta's own AI infrastructure investment. The custom chip progress matters because in-house silicon reduces Meta's dependence on Nvidia and could lower long-run AI infrastructure costs, improving the return on the billions it has been spending. Investors who were worried Meta was spending without a visible payoff got two concrete data points that the spending is producing real, deployable products.
Meta Q2 earnings report (expected late July): will confirm whether AI products are generating revenue at scale. Any Nvidia earnings or guidance updates (~late May): will show whether Meta's custom chip progress is actually reducing its GPU orders.
Full analysis · Subscribers
The deep dive (bull case, bear case, and the data point that decides which side wins), the cause-and-effect chain behind the move, plain-English explainers for every block, and the live update timeline (5 updates so far).
Aggregated reads dozens of sources in five languages and turns the day into plain-English cards like this one.
Educational analysis of public information, not investment advice. Report an error · Corrections policy
← Today's brief