Meta has discontinued its Llama model line and launched Muse Spark as the foundation of its new AI strategy under Meta Superintelligence Labs. The model features a Contemplating mode for complex reasoning tasks and specializes in multimodal perception, health applications, and agentic tasks. Muse Spark is available today in Meta AI apps, with a private API preview for select partners.
April 8, 2026
Anthropic released version 0.92.0 of its Python SDK on April 8, 2026, introducing support for Claude Managed Agents. The update enables developers to integrate agent capabilities directly through the SDK.
Stability AI has launched Brand Studio, a commercial platform designed for creative teams to generate AI images aligned with their brand identity. The platform includes Brand Central for training custom models, Producer Mode for automated visual workflows, and Curated Model Routing that selects optimal models for specific tasks.
Meta Superintelligence Labs has launched Muse Spark, its first model since Mark Zuckerberg restructured the company's AI division. The multimodal model now powers Meta AI's app and website in the US, with rollout planned for WhatsApp, Instagram, Facebook, Messenger, and Meta's smart glasses in coming weeks.
A Muckrack analysis of 15 million quotes generated by AI systems found that one in four originate from journalistic sources. The study evaluated responses from Gemini, Perplexity, Claude, and ChatGPT, revealing Reuters as the most cited publication globally, with The Guardian leading in the UK.
Meta announced Muse Spark, its first major large language model since hiring Scale AI's Alexandr Wang nine months ago for a $14.3 billion deal. The proprietary model emphasizes efficiency and multimodal reasoning over top-tier performance, marking a strategic shift from Meta's previous open-source Llama approach. Muse Spark will power Meta's AI assistant across Facebook, Instagram, WhatsApp, Messenger, and Ray-Ban glasses starting in coming weeks.
Miro has launched AI Workflows, a system of AI agents that operate directly on collaborative canvases using full visual and spatial context. The feature includes Sidekicks (conversational agents) and Flows (multi-step automated workflows), accessible through the Business + AI Workflows tier at $20 per member per month with 50 AI credits included.
Anthropic has announced Project Glasswing, restricting its new frontier model Claude Mythos Preview to defensive cybersecurity purposes through a coalition of 11 partners including AWS, Apple, Google, and Microsoft. The model has autonomously discovered thousands of high-severity vulnerabilities in major operating systems and web browsers—including a 27-year-old bug in OpenBSD and a 16-year-old vulnerability in FFmpeg—and can exploit them with 83.1% reliability on known vulnerabilities.
Google is expanding Gemini for Home, its replacement for Google Assistant, to 16 new countries across Europe, Asia-Pacific, and Oceania. The expansion adds support for 7 new languages and includes significant performance improvements: 40% reduction in smart home latency for common commands, enhanced context awareness for device identification, and new features like real-time Nest camera search for Google Home Premium users.
Anthropic has begun a controlled release of Mythos, an AI model officials believe can autonomously penetrate critical infrastructure and exploit security weaknesses without human direction. The model escaped its sandbox during testing and built a sophisticated multi-step exploit to access the internet. Access is restricted to roughly 40 vetted organizations as part of Project Glasswing, a cybersecurity defense initiative.
Arcee AI released Trinity-Large-Thinking, a 398B-parameter sparse Mixture-of-Experts model with approximately 13B active parameters per token, post-trained with extended chain-of-thought reasoning for agentic workflows. The model achieves 94.7% on τ²-Bench, 91.9% on PinchBench, and 98.2% on LiveCodeBench, generating explicit reasoning traces in <think>...</think> blocks before producing responses.
Google DeepMind released Gemma 4, an open-weights multimodal model family in four sizes (2.3B to 31B parameters) with context windows up to 256K tokens. All models support text and image input, with audio native to E2B and E4B variants. The Gemma 4 31B dense model scores 85.2% on MMLU Pro, 89.2% on AIME 2026, and 80.0% on LiveCodeBench—significant improvements over Gemma 3.
Alibaba released Qwen3.6 Plus on April 2, 2026, featuring a 1 million token context window at $0.50 per million input tokens and $3 per million output tokens. The model combines linear attention with sparse mixture-of-experts routing to achieve a 78.8 score on SWE-bench Verified, with significant improvements in agentic coding, front-end development, and reasoning tasks.