Models & Labs
Meta Launches Muse Spark as Llama Replacement
Sam Witteveenmedium confidence
Why it matters
- →The launch of Muse Spark signifies Meta's continued innovation in AI models, potentially impacting the development and deployment of language models in various applications.
Meta AI has launched Muse Spark, which serves as the internal replacement for its previous Llama models. The announcement includes performance benchmarks and a demo, showcasing the capabilities of Muse Spark compared to its predecessor. This development indicates Meta's ongoing commitment to advancing its AI model offerings. The introduction of Muse Spark may influence the competitive landscape of large language models.
Read originalThe story around this
Earlier coverage that leads up to this article, and what followed. Lines connect each piece to the closest one after it, converging here.
How we got here
- 1Google Lyria 3 AI Music Model Released
Lev Selector · February 27, 2026 · Background
What happened next
- 2Meta Launches Muse Spark AI
Matt Wolfe · May 15, 2026 · Same story
- 3Meta Releases Muse Spark Model
Lev Selector · May 15, 2026 · Same story
- 4Meta Launches Muse Spark 1.1 for AI Coding
TechCrunch AI · July 9, 2026 · Same story
- 5Meta Launches Muse Spark 1.1
AI Explained · July 10, 2026 · Same story
- 6Meta AI Launches Muse Spark Agent
Matt Wolfe · July 31, 2026 · Same story
Follow this story
Open the full story →Meta Launches Muse Spark as Llama Replacement
3 developments
- Apr 9 · Sam Witteveen
Meta Launches Muse Spark as Llama Replacement (This article) - Apr 9 · The Rundown AI
Meta Superintelligence Labs launches Muse Spark model↳ Muse Spark handles voice, text, and images but lacks strong coding skills compared to competitors. - Apr 10 · Lev Selector
Meta Releases Muse Spark Model
More in Models & Labs
Models & Labsother
llama.cpp b12340 adds CUDA 13 and ROCm 10.0
This release quietly closes the hardware gap for local inference by adding default support for CUDA 13 and ROCm 10.0 across Linux and Windows. Users with newer NVIDIA GPUs or AMD accelerators no longer need to compile from source to get performance, while Apple Silicon builds now ship KleidiAI disabled by default, shifting the focus back to standard ARM optimizations. The inclusion of Snapdragon support for Linux and Android marks a significant step toward on-device AI beyond the x86/ARM desktop ecosystem. It is less about new features and more about ensuring llama.cpp remains the universal runtime for every available accelerator.
llama.cpp Releases
Models & Labscoding
llama.cpp b12342 adds few-row MMQ for speculative decoding
Speculative decoding just got a significant speed boost on NVIDIA GPUs. This release introduces a specialized kernel for 'few-row' matrix multiplications, targeting the small batch sizes typical of verification passes in speculative decoding workflows. By optimizing these specific operations for quantized formats like Q8_0 and IQ4 variants, llama.cpp reduces latency where it matters most during token generation. The change is CUDA-only for now but leverages hardware capabilities from Ampere onwards, making local inference faster without requiring new model architectures or complex configuration changes.
llama.cpp Releases
© TechCrunch AIModels & Labsmodels
Google deploys Gemini 4 Hydrogen for faster Search
Google is quietly shifting its search infrastructure by replacing older fast models with Gemini 4 Hydrogen, a tier designed specifically for speed and cost-efficiency rather than raw reasoning power. This model completes tasks 2.5 times faster than its predecessor while beating the more powerful Argon on specific intelligence benchmarks, proving that specialized efficiency can outperform brute force in high-volume workflows. By integrating this into Search's fan-out technique, Google now breaks down complex queries into subqueries with greater precision and speed, delivering richer visual responses without the latency of heavier models. This marks a strategic pivot toward agentic execution layers within consumer search, making AI Mode feel snappier for everyday multi-part questions.
TechCrunch AI