Meta and Anthropic Launch Advanced AI Models for Complex Business Tasks

Meta and Anthropic have unveiled new artificial intelligence models designed to handle more complex and sustained workloads, marking a significant shift in the competitive landscape for AI capabilities. The announcements focus on expanding the utility of AI systems for sophisticated technical tasks, including the coordination of multiple agents and advanced software engineering. These releases highlight a broader industry trend towards making AI agents more practical for enterprise use, with an emphasis on cost efficiency, speed and the ability to execute intricate workflows.

Meta introduced Muse Ember on Monday, a smaller and more efficient model built to manage repetitive tasks performed by AI agents. The system is capable of handling tool calls, coding subtasks, information extraction and document analysis. It can also coordinate business workflows across various applications and process large collections of documents or code within a single request. The model is available through Meta’s Model API at a cost of $0.10 per million input tokens and $0.50 per million output tokens. Meta stated that internal and third-party evaluations indicate Muse Ember performs competitively against models from Anthropic and OpenAI, particularly in agent-based workflows and coding. The company also noted the model’s ability to process text, images, audio and video, positioning it for a wide range of enterprise applications.

Anthropic simultaneously launched Fable 5.2 and Mythos 5.2, which are aimed at more demanding technical assignments. Fable 5.2 can coordinate more than 1,000 AI agents simultaneously and tackle advanced software engineering projects, including hardware-related work such as controlling robots. Anthropic reported that both new models are faster than their predecessors, with Fable also being less expensive. The pricing for Fable has been reduced to $8 per million input tokens and $40 per million output tokens, down from $10 and $50 respectively for the previous version. Mythos 5.2 uses the same underlying model but features fewer safeguards for cybersecurity and life sciences applications. Access to this model remains restricted to organizations participating in Anthropic’s verification program.

The releases have also drawn attention to the safety challenges associated with increasingly capable AI systems. Anthropic stated that it subjected the new models to longer testing periods and additional evaluations, including assessments of their ability to escape digital sandbox environments. The company reported that the latest versions were less likely to breach those restrictions than their predecessors. For both companies, the announcements underscore a market focus on making AI agents practical for business use, where cost, speed and workflow execution are becoming key measures of competitiveness.