IBM Bob 自托管版正式发布：支持本地、私有云与气隙环境，代码不出内网

IBM 已经为 IBM Bob 提供了自托管部署选项：https://www.ibm.com/new/announcements/ibm-bob-expands-to-self-hosted-environments-for-sensitive-and-mission-critical-enterprise-software 并已正式上市。Bob：https://bob.ibm.com/ 是 IBM 的智能软件开发平台。它涵盖完整生命周期：理解代码、规划工作、执行更改和验证结果。新的选项允许企业在本地、私有或主权云以及隔离网络中运行 Bob。

今天可以部署吗？可以。自托管选项已向企业客户全面开放。您必须自行获取、许可并托管受支持的模型。IBM 尚未公布定价，并将购买者引导至演示请求：https://www.ibm.com/forms/mkt-mail-ibmbuy。

自托管部署包括：

Bob 自托管不捆绑模型。客户需从 IBM 支持的列表中选择模型。IBM 按部署方式对列表进行分组：

此选择决定了代码和上下文的处理位置。使用自托管模型时，代码、开发上下文和构建产物可以保留在客户环境内。

IBM 给出了一个具体示例。某银行可以在本地对核心银行应用进行推理。同一银行可以将限制较少的工作负载路由到经过批准的外部模型。开发者在两种情况下都能看到一致的核心 Bob 体验，同时平台和安全团队可以控制底层架构。

IBM 还计划扩展模型组合并增加多模型路由。这应被视为路线图，而非已发布功能。

资料来源：IBM：https://www.ibm.com/new/announcements/ibm-bob-expands-to-self-hosted-environments-for-sensitive-and-mission-critical-enterprise-software, GitLab 文档：https://docs.gitlab.com/administration/gitlab_duo_self_hosted/, Mistral Vibe：https://mistral.ai/products/vibe/, Mistral 编码栈：https://mistral.ai/news/codestral-25-08, GitHub 更新日志：https://github.blog/changelog/2026-04-07-copilot-cli-now-supports-byok-and-local-models/, GitHub 文档：https://docs.github.com/en/enterprise-server@3.22/copilot/copilot-on-ghes/about-copilot-on-ghes。

区别在于范围。GitLab 和 GitHub 将代理引入自身的 DevOps 平台。Mistral 将其代理与自己的开源权重模型配对。IBM 则专注于现代化长期存在的资产，特别是 Java、IBM i 和大型机代码，在无法接触公共互联网的环境中进行改造。

需要与我们合作以推广您的 GitHub 仓库、Hugging Face 页面、产品发布或在线研讨会等吗？请联系我们：https://forms.gle/MJjjVDPS7whH8Ngs6

Sana Hassan 是 Marktechpost 的咨询实习生，也是印度理工学院马德拉斯分校的双学位学生，热衷于将技术和人工智能应用于解决现实世界的挑战。对解决实际问题充满兴趣的他，为人工智能与现实生活解决方案的交汇处带来了新颖的视角。

Datalab 推出 OmniExtractBench，以修正提取基准中的偏差和不透明问题：https://pxllnk.co/hxplrq

以实践者为先的 AI/ML 新闻与分析，每月有超过 100 万开发者和研究人员阅读。

IBM has made a self-hosted deployment option for IBM Bob：https://www.ibm.com/new/announcements/ibm-bob-expands-to-self-hosted-environments-for-sensitive-and-mission-critical-enterprise-software generally available. Bob：https://bob.ibm.com/ is IBM’s agentic software development platform. It covers the full lifecycle: understanding code, planning work, executing changes and validating results. The new option lets enterprises run Bob on premises, in private or sovereign clouds, and in air-gapped networks.

Is it deployable today? Yes. The self-hosted option is generally available to enterprise customers. You must source, license and host a supported model yourself. IBM has not published pricing and routes buyers to a demo request：https://www.ibm.com/forms/mkt-mail-ibmbuy.

The self-hosted deployment includes:

Bob self-hosted does not bundle a model. Customers bring one from IBM’s supported list. IBM groups the list by deployment method:

This choice determines where code and context get processed. With a self-hosted model, code, development context and build artifacts can remain inside the customer environment.

IBM gives a concrete example. A bank could run on-premises inference for core banking applications. The same bank could route a less restricted workload to an approved external model. Developers see a consistent core Bob experience in both cases, while platform and security teams control the architecture underneath.

IBM also plans to expand the model portfolio and add multi-model routing. Treat that as roadmap, not a shipped feature.

Sources: IBM：https://www.ibm.com/new/announcements/ibm-bob-expands-to-self-hosted-environments-for-sensitive-and-mission-critical-enterprise-software, GitLab docs：https://docs.gitlab.com/administration/gitlab_duo_self_hosted/, Mistral Vibe：https://mistral.ai/products/vibe/, Mistral coding stack：https://mistral.ai/news/codestral-25-08, GitHub changelog：https://github.blog/changelog/2026-04-07-copilot-cli-now-supports-byok-and-local-models/, GitHub Docs：https://docs.github.com/en/enterprise-server@3.22/copilot/copilot-on-ghes/about-copilot-on-ghes.

The differentiator is scope. GitLab and GitHub bring agents to their own DevOps platforms. Mistral pairs its agent with its own open-weight models. IBM targets modernization of long-lived estates, especially Java, IBM i and mainframe code, inside environments that cannot touch the public internet.

Need to partner with us for promoting your GitHub Repo OR Hugging Face Page OR Product Release OR Webinar etc.? Connect with us ：https://forms.gle/MJjjVDPS7whH8Ngs6

Sana Hassan, a consulting intern at Marktechpost and dual-degree student at IIT Madras, is passionate about applying technology and AI to address real-world challenges. With a keen interest in solving practical problems, he brings a fresh perspective to the intersection of AI and real-life solutions.

Datalab Introduces OmniExtractBench to Fix Bias and Opacity in Extraction Benchmarks ：https://pxllnk.co/hxplrq

Practitioner-first AI/ML news and analysis, read by 1M+ developers and researchers every month.