Meta and SpaceX are racing into the developer market with new A.I. models, Grok 4.5 and Muse Spark 1.1, respectively, trailing OpenAI and Anthropic by leaning on coding and agentic capabilities with lower token pricing. Meta unveiled Muse Spark 1.1 on the 9th, featuring agentic coding that can manage an entire software project, and marking the first time Meta has offered a commercial A.I. application programming interface (A.P.I.) to outside developers. Previously, Meta made Llama models available only as open-weight models, with third parties supplying access; Muse Spark 1.1 represents Meta's first direct entry into the A.I. A.P.I. business. Meta has poured hundreds of billions of dollars into A.I. development but generated little direct revenue from models, as its core business remains advertising. Meta's stock rose 4.6 percent on the 9th after sliding on a Reuters report about producing its own chips starting in September. Muse Spark 1.1 was built by Meta Superintelligence Labs under Alexander Wang, who joined after Meta acquired a 49 percent stake in Scale A.I. for $14.3 billion last June. Meta rebuilt its infrastructure for Muse Spark, a distinct lineage from Llama, and may keep releasing smaller Llama 4 models like Scout (109 billion parameters) and Maverick (400 billion parameters), while restricting Behemoth (2 trillion parameters) to research.