Meta Expands AI Coding Capabilities with Muse Spark 1.1
Meta has released Muse Spark 1.1, an upgraded AI model designed to challenge industry leaders like OpenAI and Google by integrating directly into developer workflows. Available through a new API, the model focuses on complex bug detection and multimodal processing, marking a significant step in Meta’s aggressive pursuit of generative AI parity.
The update arrives just months after the company's initial re-entry into the AI space with the original Muse Spark. Developers can now utilize the model for sophisticated agentic workflows and multi-agent systems, capabilities that Meta claims were refined based on extensive user feedback. The model processes images, videos, and documents natively, signaling a move toward more versatile coding assistants.
Accessibility is a core component of this rollout. While the model is currently available via the Meta AI website and app in a specialized "Thinking" mode, the new Meta Model API offers a gateway for external integration. US-based developers are receiving $20 in free credits to test the public preview. This deployment follows the recent release of Muse Image, which faced scrutiny over its use of Instagram content for training data. Meta continues to pivot its massive infrastructure investments toward these tools, aiming to embed its AI technology deeper into its ecosystem, ranging from WhatsApp chatbots to its latest smart glasses.
Comments (0)
No comments yet. Be the first!