
Berryxia.AI
8.2K posts

Berryxia.AI
@berryxia
Building AI tools AI System Prompt Love Design & Coding & Share Prompt! 📮:[email protected]



兄弟们,这个笔记本电脑多模态模型有点东西! PrismML 今天正式发布了 Bonsai 27B,这是目前全球首个能在手机上本地运行的 27B 级大模型。 传统 27B 模型在 FP16 下需要约 54GB 内存,即使激进量化到 4-bit 也仍有 18GB 左右,手机和大多数笔记本根本无法承载。 而 Bonsai 27B 通过端到端极致低比特技术,成功将模型压缩到: - 1-bit 版本:仅 3.9GB,可直接在 iPhone 17 Pro 上运行 - Ternary 版本:5.9GB,适合笔记本高性能使用 两者均为多模态模型,内置 4-bit 视觉塔,支持图片、截图、文档等输入,同时保留 262K 超长上下文和投机解码加速能力。 更惊人的是其智能密度达到了 0.53 per GB,比全精度基线高出约 10 倍,比现有最佳低比特方案高出 2.7 倍。 性能保留方面,在 15 个基准测试(thinking mode)中,Ternary 版平均保留 95%,1-bit 版保留 90%,数学、编码和工具调用能力尤为接近原版 Qwen3.6 27B。 这带来的最大价值在于真正实现了设备原生 Agentic AI。 用户可以在手机或笔记本上本地跑多步推理、结构化工具调用、长期 agentic 工作流,完全离线且数据不离开设备。 无论是处理私人文档、持续研究任务,还是构建隐私优先的智能体,都不再依赖云端。 目前模型已全部开源(Apache 2.0 协议),开发者可直接在 Hugging Face 下载 GGUF 格式模型,并通过官方提供的 Bonsai-demo 一键运行(支持 llama.cpp 和 MLX)。 官网也开放了限时免费开发者预览 API 供快速测试。 从云端大模型到口袋里的 27B 智能,Bonsai 27B 真正把智能密度推向了新的高度,也为本地 Agent 和混合部署打开了全新的可能性。 官网:prismml.com/news/bonsai-27b 模型集合:huggingface.co/collections/pr… 一键 Demo:github.com/PrismML-Eng/Bo…

Today, we’re announcing Bonsai 27B: the first 27B-class model to run on a phone. Bonsai 27B is the new multimodal flagship of the Bonsai family. Based on Qwen3.6 27B, it brings a new capability tier to local AI: multi-step reasoning, structured tool use, long-context workflows, and coherent agentic loops. Until now, models in this class have been impractical to deploy locally. A 27B model occupies roughly 54 GB in 16-bit precision, and even a strong 4-bit build is around 18GB - too large for a phone and for most laptops. Bonsai 27B changes that. It comes in two variants: • Ternary Bonsai 27B: 5.9 GB, 1.71 effective bits per weight, optimized for laptop-class quality. • 1-bit Bonsai 27B: 3.9 GB, 1.125 effective bits per weight, optimized for phone-class footprint. Everything is open-sourced today under the Apache 2.0 license.

🤗 MOSS-VL-Realtime is now open source on @huggingface . The 11B model family supports text, single and multiple images, single and multiple videos, and interleaved visual-text inputs in Chinese and English.@MosiAI_Official Highlights: 🏗️ Cross-Attention architecture separating visual encoding from language reasoning 🧭 XRoPE for unified temporal-spatial positioning 🧩 Unified conversation templates for offline, streaming, and real-time interaction 🧠 256K-token context window 📜 Apache-2.0 license MOSS-VL-Realtime continues processing new frames while generating a response, allowing it to revise or interrupt that response as the scene evolves—or remain silent when more evidence is needed. Thank you @sgl_project @lmsysorg for day-0 support! 🚀 Huggingface: huggingface.co/OpenMOSS-Team/… Github: github.com/OpenMOSS/MOSS-… Technical blog: openmoss.github.io/MOSS-VL Join the community: discord.gg/SmVQHGffZU 👇

The new PrismML’s Bonsai 27B model running on iPhone 17 Pro Bonsai 27B is the first 27B-class model to run on a phone and it uses less than 5GB of memory Pushing the limits on what can run a phone

New tools I keep coming back to. Save this 👇 - motionsites.ai - AI prompts for animated websites - originkit.dev - free animated component library - text-effects.colorion.co - pure CSS text animations - kinetics.colorion.co - spring physics UI motion - opentui.com - build terminal UIs with TypeScript - namethatui.com - name any UI element instantly - shapefactory.co - logo, color & gradient tools - valessa.riotters.com - 3D product visualizer in browser - github.com/MengTo/Skills - Claude skills for web design What am I missing?

New feature on Wan Video: **Music to Dance** 💃 Upload a character, add a song, and let Wan Video generate a dance video synchronized to the rhythm. Available dance styles: • Street • Tap • Latin • K-Pop • Chinese Classical From beat to movement, your character performs with the music.

GPT 5.6 Sol can one-shot convert an arXiv paper into an interactive Marimo notebook! Great for papers best understood hands on (lots of fun examples in interpretability, inference engineering, agent harnesses, benchmarking, and more) Play around with the notebook, inspect the code, or try the same workflow with your own agents below




We're introducing Claude for Teachers: free access to premium Claude capabilities for verified K-12 educators in the US, with a library of teaching skills and a direct connection to evidence-based curricula, mapped to academic standards in all 50 states. claude.com/solutions/teac…










