Meta Platforms is preparing to release an improved version of its Muse Spark AI model, featuring more sophisticated coding capabilities that reportedly rival OpenAI’s GPT-5.5. The company’s chief AI officer confirmed the rollout is expected soon, highlighting significant enhancements in both coding performance and AI agent integration.
- New Muse Spark model rivals GPT-5.5 on key coding benchmarks.
- Enhanced AI agent capabilities included with updated model.
- Model demands significantly more computing power than predecessor.
What happened
Meta Platforms announced it will soon release an updated version of its Muse Spark AI model, designed to significantly improve its coding capabilities. The announcement came from Alexandr Wang, Meta’s chief AI officer, who indicated the enhanced model is close to launch. The update arrives shortly after reports suggested the new Muse Spark variant can compete with OpenAI’s GPT-5.5 on several important AI coding benchmarks.
The original Muse Spark scored 52.5% on the SWE-Bench Pro coding test, while GPT-5.5 scored 58.6%. Meta’s upcoming model shows a clear improvement over the prior version in coding proficiency and the ability to power AI agents, a feature used to enhance response quality. The new algorithm utilizes more computational resources, reflecting a trade-off between performance and infrastructure demands.
Why it matters
Meta’s move to improve Muse Spark’s coding skills highlights its ambition to compete directly with market-leading AI developers like OpenAI and Anthropic. Coding benchmarks such as SWE-Bench Pro and Terminal-Bench 2.0 are widely regarded metrics for evaluating AI programming capabilities. The model’s advancement could position Meta as a stronger contender in the AI development space, especially if integrated within external developer tools or offered as an AI infrastructure service.
Additionally, improving AI agent functionality through features such as Muse Spark’s 'contemplating mode' indicates a focus on increasing the practical utility of AI in real-world applications. The push to enhance coding performance alongside agent intelligence supports Meta’s broader strategy to expand offerings beyond social media into AI-powered software solutions across technical and nontechnical domains.
What to watch next
Industry observers will be looking for Meta’s official release timeline and whether the company will open access to external developers or enterprise customers. Bloomberg recently reported that Meta is considering launching AI infrastructure services, and the new model’s computational demands may influence how these services are structured—whether as raw computing capacity or fully hosted AI models.
Competitive dynamics with Anthropic’s Claude Code and OpenAI’s GPT line remain critical, especially since Meta aims to surpass current coding leaders. Future announcements are expected to reveal how Meta plans to integrate Muse Spark’s capabilities into developer ecosystems and productivity tools, which could accelerate AI adoption across sectors including healthcare, finance, and automation.