What happened
Meta released Muse Spark 1.3, which per AAII is now the #3 model in the world. Mark Zuckerberg announced it is rolling out today with what he calls frontier performance 'almost too cheap to meter,' and promised open weights releases coming soon.
Why it matters
The model is positioned as Meta's biggest jump yet on coding and agentic work, and the pricing model is over 90% cheaper if you opt in to training. This makes frontier-level AI dramatically more accessible, potentially reshaping the competitive landscape.
What to watch
The reported long-context result of 98.1% on MRCR 512k–1m is notable. Some commenters question whether this means Muse Spark has solved 'context rot' at million-token scale, which would be a significant technical achievement.
Ask the AI about this article →
Summaries like this, in your inbox every morning.
For example, today's edition would include:
AI-summarized, only the topics you pick — one digest a day via Email, Slack, or Discord.
Free · 30 seconds with Google · unsubscribe anytimeWhat is AIToday? →
Ask AI anything about this article. Q&As are published on this page for other readers too.
Lightbits Labs announced general availability of Inferra, a software engine that manages KV cache data across…
Databricks expanded its Adaptive Instructed-Retriever search model, a retrieval building block for its Genie C…
Sequoia co-led a $25 million Series A for Cymphony, valuing it above $100 million, to secure AI agents in ente…
Connor Leahy, U.S. Executive Director of nonprofit ControlAI, told TechCrunch's Equity podcast that AI compani…
Jacob Coxon, who spent three years on pre-training at OpenAI and Anthropic, resigned Tuesday, accusing the fir…
Instinct, the AI assistant valued at $2.5 billion, is rolling out its own email addresses so it can create and…
