Thinking Machines 推出 26B 参数的 Inkling 2 Mini,已上线 Agent Arena
Thinking Machines 这次把模型压到了 26B,笔记本就能跑,还能在 Agent Arena 里用搜索和终端干活。之前嫌 594B 的 Inkling 2 Small 太大的朋友可以试试这个小的,榜单位置就等你去测了。
Thinking Machines 发布 Inkling 2 Mini,采用 looped transformer 架构,参数量仅 26B,比此前 594B 的 Inkling 2 Small 小了超过 20 倍,可在笔记本电脑上运行。该模型在 AA Index 上取得 20.1 分,官方称其足以胜任智能体任务。
模型现已进入 Agent Arena 排行榜,该平台基于全球用户提交的数百万条真实长程智能体任务评测,模型可调用网页搜索、文件系统和终端工具完成复杂工作流,并使用因果追踪方法衡量相对表现。Inkling 2 Mini 同时上线了 Code Arena 的 WebDev 赛道和 Text Arena,具体成绩尚未公布。
在模型普遍向百亿至千亿参数扩张的背景下,主打小尺寸但保留智能体能力的 looped transformer 是一条差异化路线。对于想在本地硬件上跑智能体工作流的开发者,26B 量级提供了除云端 API 之外的另一选择。
Inkling 2 Mini by @thinkymachines is now in the Arena. Head to Agent Arena to test out the small 26B model, scores coming soon! In Agent Arena, we measure models on millions of real-world, long-horizon agentic tasks from a global community of users. Models can access web search, filesystem, and terminal tools to complete complex workflows. The leaderboard measures model performance on outcomes relative to the average model using a causal tracing methodology. Inkling 2 Mini is also available in the Code Arena: WebDev and in the Text Arena. Thinking Machines @thinkymachines Some of you thought that Inkling 2 Small was too big at 594B parameters. We heard you: Inkling 2 Mini is a looped transformer with just 26B parameters, small enough to run on your laptop, powerful enough for agentic work with a score of 20.1 on AA Index. thinkingmachines.ai/news/inkling-2… 🔗 View Quoted Tweet📊 4 ⚡
This report is categorized under AITOP's "Models" desk, focusing on US frontier AI labs and Silicon Valley technology leaders. Within our cross-regional analytical framework, it directly maps to AI developer toolchains, coding agents, and engineering workflows and US-China foundation model benchmarks and commercialization comparison. Key entities and technical topics under continuous tracking include Inkling 2 Mini, Thinking Machines, Agent Arena, open-source models.
Source attribution: curated from lmarena.ai (@lmarena_ai).