跳到正文
只看模型#开源生态
  1. Rohan Paul78

    Recursive 发布首个模型 Flint，这是一个 40B MoE 本地编码模型，采用 Apache 2.0 许可，权重已在 Hugging Face 开放。

    引用Richard Socher@RichardSocher

    Introducing Flint, @Recursive_SI’s first model: a fast local coding model built with our automated research system, the Eureka Machine. • Best-in-class fast local coding • Up to 2x faster than comparable models: 125+ tokens/sec on a MacBook Pro, 150+ on RTX Spark • Built from scratch in under 4 months Private, unmetered intelligence on your own machine. Open weights available now on @huggingface, run it locally on your favorite framework or try it for free on @opencode or @pidotdev.

    推荐理由：材料给出了本地编码模型的速度与质量对比数据，可据此判断本地部署在速度与质量之间的取舍。

  1. Hugging Face78

    Reflection 宣布 Beam 已在多平台上线，以 Apache 2.0 协议开放权重，并与 40 多家生态伙伴合作。此次发布包含权重、tokenizer 和参考实现，BF16、FP8、NVFP4 权重及 DFlash2 drafter，以及训练 Beam 所用的开源工具 Mirror CLI。Hugging Face 转发该消息并预告 Reflection 团队直播讲解 Beam 的构建过程。

    引用Reflection@reflection_ai

    Beam is now available across platforms, with open weights under the Apache 2.0 license - in collaboration with an ecosystem of 40+ partners. We're releasing: - Weights, tokenizer, and reference implementation - BF16, FP8 and NVFP4 weights and a DFlash2 drafter - The open source harness we built and trained Beam on: Mirror CLI Weights and model card on Hugging Face: https://huggingface.co/reflection

    推荐理由：Reflection 公开 Beam 的权重、量化版本与训练工具 Mirror CLI，可据此了解其开放程度与复现路径。

  1. Qwen70

    Qwen 发布 Qwen-Image-2.1-Turbo，在同样的 7B 视觉生成架构上做加速，仅需 8 步去噪即可生成和编辑图像，权重已在 ModelScope 和 Hugging Face 开放。

    引用Qwen@Alibaba_Qwen

    Meet Qwen-Image-2.1, the most balanced and cost-effective image generation model in the Qwen-Image series! Now open weights! 🎨 A unified model for both generation and editing, delivering top-tier quality in a lightweight package. Highlights: 👀 - Compact & exceptionally fast: A lightweight 7B architecture that outperforms most closed-source models, with drastically accelerated inference for multi-image inputs. - Native transparency: Natively generates and edits RGBA layers, enabling seamless compositing and text editing within transparent images. - Versatile, high-fidelity editing: Supports up to 10 reference images and precise local control while preserving strict fidelity for portraits and products. - Broad coverage & stunning aesthetics: Excels at panoramas, infographics, and virtual try-ons, delivering realistic textures and elegant typography. Start to create your next masterpiece with Qwen-Image-2.1! 🖼️ - Blog: https://qwen.ai/blog?id=qwen-image-2.1 - GitHub: https://github.com/QwenLM/Qwen-Image-2.1 - Model Scope: https://www.modelscope.cn/models/Qwen/Qwen-Image-2.1 - Hugging Face: https://huggingface.co/Qwen/Qwen-Image-2.1

    推荐理由：官方给出 Turbo 加速版在 8 步去噪下仍能生成 2K 图像，并同步开放权重与 API，便于判断落地成本。

  1. Google DeepMind72

    Google DeepMind 发布 EmbeddingGemma 2 端侧多模态嵌入模型

    Google DeepMind 发布 EmbeddingGemma 2，这是基于 Gemma 4 架构、采用 Apache 2.0 许可的 740M 参数嵌入模型，可将文本、代码、图像、音频和视频映射到统一嵌入空间。

    推荐理由：原文给出参数量、内存占用和基准提升，可据此判断端侧多模态检索的落地条件。

  1. Qwen75

    阿里 Qwen 发布 Qwen-Image-2.1 并开放权重，这是 Qwen-Image 系列中兼顾效果与成本的图像生成模型，统一支持生成与编辑。该模型采用 7B 轻量架构，官方称其性能超过多数闭源模型，并大幅加速多图输入推理。它可原生生成和编辑 RGBA 透明图层，支持最多 10 张参考图与局部精确控制，覆盖全景图、信息图和虚拟试穿等场景。

    推荐理由：官方给出 7B 轻量架构、原生 RGBA 透明图层与最多 10 张参考图编辑等能力，可据此判断开源图像模型的实际可用边界。

  1. Mira Murati78

    Thinking Machines 发布首个模型 Inkling，从零训练，权重开放，即日起可在 Tinker 上微调。据其介绍，Inkling 能跨文本、图像和音频模态进行高效推理，完整权重已公开，并可在 Inkling Playground 中试用。

    引用Thinking Machines@thinkymachines

    Today, we are introducing Inkling. Inkling reasons efficiently across text, image, and audio modalities. We are making the full weights available. https://thinkingmachines.ai/news/introducing-inkling/ Available today for fine-tuning on Tinker. Play with it in the Inkling Playground. 🧵

    推荐理由：Thinking Machines 首款模型从零训练并开放权重，读者可据此了解其多模态推理定位与微调入口。

已经到底了