Meta Superintelligence Labs Releases Muse Spark 1.1: A Multimodal Reasoning Model for Agentic Tasks on Meta Model API
Meta Superintelligence Labs launched Muse Spark 1.1, a multimodal reasoning model for agentic tasks with a 1 million token context window and zero-shot tool generalization. The model is now available via the Meta Model API public preview.
- Muse Spark 1.1 introduces a 1 million token context window, enabling processing of vast amounts of multimodal data.
- The model supports zero-shot generalization to new tools and MCP servers, enhancing adaptability.
- Multi-agent delegation allows parallel subagents to collaborate on complex tasks.
- Meta Model API public preview provides developers with direct access to Muse Spark 1.1.
Meta Superintelligence Labs has introduced Muse Spark 1.1, a multimodal reasoning model designed specifically for agentic tasks. The model features a groundbreaking 1 million token context window, enabling it to process and retain vast amounts of information simultaneously. Additionally, Muse Spark 1.1 supports zero-shot generalization to new tools and MCP servers, allowing it to adapt to unfamiliar environments without prior training. The model also introduces multi-agent delegation capabilities, enabling parallel subagents to work together efficiently.
The release coincides with the public preview of the Meta Model API, which provides developers with direct access to Muse Spark 1.1 and other Meta models. According to Meta's internal benchmarks, Muse Spark 1.1 leads in tool use performance but trails behind competitors like Opus 4.8 and GPT-5.5 in coding tasks. This positions the model as a strong contender in the growing field of agentic AI, where multimodal reasoning and tool generalization are becoming increasingly critical.
Muse Spark 1.1 offers a powerful new tool for building agentic AI systems with advanced multimodal reasoning and tool generalization.
Companies leveraging agentic AI can benefit from Muse Spark 1.1's capabilities to enhance automation and decision-making processes.
Meta's expansion into agentic AI with Muse Spark 1.1 signals growing investment opportunities in multimodal and tool-use models.
Muse Spark 1.1 represents a significant step forward in AI agent capabilities, particularly in handling complex, multimodal tasks.
- Agentic tasks
- AI tasks that involve autonomous decision-making, tool use, and interaction with external systems or environments.
- MCP servers
- Model Context Protocol servers that enable AI models to interact with external tools and APIs.
AI reveals hidden San Andreas Fault movements - Phys.org
Safety and innovation are complementary for the future of agentic AI - the University of South Florida
The AI Cheating Tsunami Hits the Ivy League - National Review
AI ResearchAnthropic found a hidden space where Claude puzzles over concepts
AI system developed by UC Irvine physicists helps explain why neutrinos have mass - UC Irvine News
BusinessFidji Simo steps down from OpenAI’s no. 2 role
OpenAI’s chief operating officer Fidji Simo is leaving her full‑time role after an extended medical leave, creating a leadership gap as the company prepares for a potential IPO and competes with Anthropic in enterprise AI.
BusinessFidji Simo steps down from leading OpenAI’s AGI work due to illness
OpenAI’s AGI chief Fidji Simo is leaving her full-time role to become a part-time advisor, citing health concerns after a medical leave earlier this year.
BusinessOpenAI’s CEO of AGI Deployment Fidji Simo Is Stepping Down
OpenAI’s CEO of AGI Deployment, Fidji Simo, is stepping down from her role after taking significant medical leave. She will remain as a part-time advisor.
LLMOpenAI launches its new family of models with GPT-5.6
OpenAI introduced GPT-5.6, the latest model in its new family, promising enhancements in language understanding and cybersecurity applications.
ProgrammingReturn on Attention: Why AI Code Reviews Are Wearing Us Out
AI-driven code reviews are increasing PR volume without improving quality, leading to developer burnout and hidden communication gaps.
BusinessAn AI agent startup just let its agent run its $100 million fundraise
Lyzr, an enterprise AI agent startup, autonomously managed its $100 million funding round using its own AI agent, demonstrating the product's real-world viability.