Observed Signal · Apr 8, 2026 · Technical Release · Source: Manager Magazin · Impact: 4/5 · Sentiment: Positive
Meta Launches Muse Spark to Replace Llama Models
Meta announced Muse, a new family of AI models, and unveiled Muse Spark — the first and strongest model from the Meta Superintelligence Labs. Muse Spark will power a redesigned Meta AI with improved multimodal understanding, parallel “subagent” reasoning, a shopping mode, and the ability to build experiences (e.g., games) directly from prompts. Meta said Muse Spark will roll out in the coming weeks on WhatsApp, Instagram, Facebook, Messenger and Meta AI Glasses, starting in the U.S. and expanding to other regions. API access is initially limited to select partners; Meta said it hopes to offer open‑source models in the future. The article also notes recent team moves tied to Meta’s AI effort — including Alexandr Wang as Chief AI Officer after Scale AI’s acquisition and Meta’s purchase of Moltbook, bringing Moltbook CEO Matt Schlicht and organisation lead Ben Parr into the Superintelligence Labs.
Major platform (Meta) released a new foundation model that will power conversational experiences across core social apps and devices, affecting conversational UX, potential ad placements in chat/assistant contexts, and competition in AI capabilities.
Track Meta Signals & Market Shifts in Real-Time
Polaris7 autonomous intelligence agents track regulatory filings, primary sources, executive changes, and deal flow 24/7. Create your Polaris7 workspace to monitor these entities.
Key Takeaways & Evidence Grounding
- Meta introduced Muse, a new family of AI models; Muse Spark is the first and most capable model.
- Muse Spark will power a redesigned Meta AI and add multimodal understanding, parallel subagents, a shopping mode, and prompt-driven building (e.g., games).
- Muse Spark will be rolled out in the coming weeks on WhatsApp, Instagram, Facebook, Messenger and Meta AI Glasses, starting in the United States.
- API access to Muse Spark is being provided to selected partners; Meta expressed intent to offer open‑source models in the future.
- Meta acquired Moltbook; Moltbook CEO Matt Schlicht and organisational lead Ben Parr joined the Meta Superintelligence Labs; Alexandr Wang is noted as Chief AI Officer.
Connected Companies & Entities
Ontology Mapping & Concepts
Related Market Signals & Shifts
Recent verified developments and strategic activity across this market segment.
Meta AI App Hits No. 5 After Muse Spark Launch
Meta launched Muse Spark, a new multimodal AI model, and updated its Meta AI app and website. Appfigures data show the Meta AI app jumped from No. 57 to No. 5 on the U.S. App Store immediately after the Muse Spark release, suggesting a large increase in installs. Appfigures reports 60.5 million total installs across App Store and Google Play (25 million in 2026) and a 138% download increase over the app’s first five months. Muse Spark accepts voice, text and image inputs, can spin up multiple subagents, and is designed to help with tasks such as health education, complex reasoning, and visual coding. Meta plans to roll the model out to WhatsApp, Instagram, Facebook, Messenger and its AI glasses in the coming weeks. Meta positions Muse Spark as an upgrade to Llama 4 and a step toward competing with OpenAI, Anthropic and Google.
Meta launches Muse Spark 1.1 for agentic coding
Meta publicly launched Muse Spark 1.1, a multimodal AI model aimed at agentic coding and enterprise automation. Spark 1.1 can perform multi-step reasoning, manage digital workflows, fix bugs and assist with large code migrations, according to Meta. Reuters reported Meta will price the model at $1.25 per million input tokens and $4.25 per million output tokens, putting it roughly in line with offerings from Anthropic and OpenAI. CEO Mark Zuckerberg posted about the release on X, calling Spark strong on agentic performance and computer/tool use and indicating more model releases are forthcoming. The launch positions Meta as a late but competitive entrant in an intensifying market of coding-capable foundational models.
Meta Launches Muse Spark 1.3 to Rival GPT-5.6 Sol
Meta has released Muse Spark 1.3, the latest iteration of its AI model family, designed for longer agentic workflows, complex coding tasks, and reliable processing of extensive instructions. The model improves collaboration by maintaining context in long chats, distinguishing multiple tasks, and asking for clarification. It also enhances safety against prompt injection and irreversible actions. Reportedly, Muse Spark 1.3 uses about 20% fewer tool calls and 25% fewer tokens than its predecessor, with a more compact coding style. In Meta's benchmarks, it outperforms rivals GPT-5.6 Sol and Opus 5 on the long-context MRCR test (98.5 vs 91.5 points) and several coding benchmarks, though it trails in others like DeepSearchQA. The model is available in Muse Code and via the Meta Model API, with a Max-Reasoning mode and an Open-Weights version announced.
Receive all future signals for Meta & competitors in your briefing
Never miss material M&A deals, leadership changes, account wins, or product rollouts. Polaris7 monitors 20,000+ primary sources 24/7, delivering verified telemetry directly into your executive briefing.
