Thinking Machines Lab Drops Its First Model
Thinking Machines Lab Drops Its First Model
Thinking Machines Lab 发布其首款模型
Thinking Machines Lab, an artificial intelligence company started by exiles from OpenAI, has released its first model, called Inkling. The startup’s new model is open-weight, which means that researchers and startups will be able to download and modify it.
由 OpenAI 前员工创立的人工智能公司 Thinking Machines Lab 发布了其首款模型，名为 Inkling。这家初创公司的新模型采用开放权重（open-weight）模式，这意味着研究人员和初创企业可以下载并对其进行修改。
In a blog post, the company says Inkling was trained from scratch to make sense of audio and video input as well as text. It says that while Inkling isn’t the best model on popular benchmarks, it performs well at many tasks, and is capable of advanced reasoning and coding. Like many open-weight models, Inkling is relatively large—975 billion parameters—and needs to run on a cluster of specialized chips.
该公司在一篇博客文章中表示，Inkling 是从零开始训练的，旨在理解音频、视频以及文本输入。公司称，虽然 Inkling 在主流基准测试中并非表现最强的模型，但在许多任务中表现出色，并具备高级推理和编程能力。与许多开放权重模型一样，Inkling 的规模相对较大，拥有 9750 亿个参数，需要在专门的芯片集群上运行。
In a sign of how AI models are increasingly being used to build AI, the lab used Inkling to fine-tune and improve itself. The training process also surfaced an interesting phenomenon: Like other models, Inkling usually provides a natural language explanation for its complex reasoning. According to the company’s blog post, in order to perform more efficiently, “the chain of thought became more concise over time, dropping grammatical overhead while remaining comprehensible and leaving the final response unaffected.”
作为一个 AI 模型正越来越多地被用于构建 AI 的标志，该实验室利用 Inkling 对其自身进行了微调和改进。训练过程还揭示了一个有趣的现象：与其他模型一样，Inkling 通常会为其复杂的推理过程提供自然语言解释。据公司博客称，为了提高运行效率，“思维链（chain of thought）随着时间的推移变得更加简洁，在保持可理解性且不影响最终响应的前提下，去除了冗余的语法开销。”
The release could help Thinking Machines establish itself as a legitimate player in the frenetic and big-spending AI race. Open-source models have proven popular because they’re cheaper to run than closed models, which can typically only be accessed for a fee. Open-source models can also be more easily modified for different tasks. The best open-weight models currently come from China, but Thinking Machines says Inkling offers a level of performance similar to those models.
此次发布有助于 Thinking Machines 在竞争激烈且投入巨大的 AI 竞赛中确立其作为重要参与者的地位。开源模型之所以广受欢迎，是因为它们比通常需要付费才能访问的闭源模型运行成本更低。此外，开源模型也更容易针对不同任务进行修改。目前，最优秀的开放权重模型多来自中国，但 Thinking Machines 表示，Inkling 的性能水平与这些模型相当。
The release of an open-weight model fits with a vision for AI that Thinking Machines laid out in a recent blog post. The company said the technology shouldn’t be controlled by just a few companies and should be decentralized so that more people can build their own models with their own data.
发布开放权重模型符合 Thinking Machines 在近期博客中提出的 AI 愿景。该公司表示，这项技术不应仅由少数几家公司控制，而应实现去中心化，以便更多人能够利用自己的数据构建属于自己的模型。
Thinking Machines was founded in February 2025 by several big-name executives and researchers from OpenAI, including Mira Murati, who served as CTO (and briefly CEO) of OpenAI; John Schulman, a cofounder of OpenAI who played a key role in developing ChatGPT; and Lilian Weng, a former VP at OpenAI who led work on safety and robotics.
Thinking Machines 由多位来自 OpenAI 的知名高管和研究人员于 2025 年 2 月创立，其中包括曾担任 OpenAI 首席技术官（及短暂担任首席执行官）的 Mira Murati；在 ChatGPT 开发中发挥关键作用的 OpenAI 联合创始人 John Schulman；以及曾领导 OpenAI 安全与机器人研究的前副总裁 Lilian Weng。
The startup received the largest seed funding round in history, which valued it at $12 billion out of the gate. Previously, the company released Tinker, a tool for fine-tuning models, showcased a tool that enables natural voice interactions, and published machine-learning research.
这家初创公司获得了历史上规模最大的种子轮融资，成立之初估值便达到 120 亿美元。此前，该公司曾发布过用于模型微调的工具 Tinker，展示过一种支持自然语音交互的工具，并发表过机器学习相关的研究成果。
OpenAI may have kick-started the AI boom with ChatGPT, but defector-led companies like Thinking Machines and Anthropic have muscled into the space. Anthropic recently filed for an IPO, which could value the company at more than a trillion dollars. Its model Claude has proven popular with many businesses, especially for its coding skills.
OpenAI 或许凭借 ChatGPT 引发了 AI 热潮，但像 Thinking Machines 和 Anthropic 这样由前员工领导的公司已经强势切入该领域。Anthropic 最近提交了 IPO 申请，其估值可能超过一万亿美元。其模型 Claude 因出色的编程能力，已受到许多企业的青睐。
Update 07/15/2026 6:09pm ET: This story has been updated to clarify a quote about the model’s training process.
更新于美东时间 2026 年 7 月 15 日下午 6:09：本文已更新，以澄清关于模型训练过程的引述。