Meta releases technical report for Muse Glimmer 30B, an agentic model for consumer-grade hardware
English text is being prepared.
Meta has released the technical report for Muse Glimmer, a 30B open-weight model aimed at agentic tasks on consumer-grade hardware; it runs at 4-bit precision on a single consumer-grade GPU and leads open models of the same scale on tool calling and multi-step planning. The relayer adds that the model surpasses Qwen3.6-27B and Gemma4-31B on some metrics, is about 17GB after 4-bit quantization, targets devices with 24GB of VRAM, and uses DFlash speculative decoding to reach about 233 tokens/s on an RTX 5090, roughly 3.1 times faster than ordinary decoding.
Why read this
AI score 75
Meta has published the technical report for its 30B open-weight agentic model, with quantization and speculative decoding providing concrete figures for running it on a single consumer-grade GPU.
Topics
View original postToday we published a technical report on Muse Glimmer, our 30B open-weight model built for agentic tasks on consumer hardware. It runs at practical speeds on a Mac or PC with a single consumer GPU at 4-bit precision, and leads comparably-sized open models on tool use and multi-step planning. Read about the model architecture, distillation recipe, quantization and more: https://t.co/H7n9qeDhxn
View original
Today we published a technical report on Muse Glimmer, our 30B open-weight model built for agentic tasks on consumer hardware. It runs at practical speeds on a Mac or PC with a single consumer GPU at 4-bit precision, and leads comparably-sized open models on tool use and multi-step planning. Read about the model architecture, distillation recipe, quantization and more: https://t.co/H7n9qeDhxn
Body
Meta 正在把具备较强自主执行能力的 AI Agent，从云端大模型压缩到可以在个人电脑上持续运行的 Muse Glimmer 30B。
部分指标超过 Qwen3.6-27B 和 Gemma4-31B。
通过 4-bit 量化，将模型压缩到约 17GB，目标是在 24GB 显存设备上运行。
采用 DFlash 推测解码，在 RTX 5090 上实现约 233 tokens/s，相比普通解码提速约 3.1 倍。
Meta 最近牛逼克拉斯！