Hermes Agent vs OpenClaw: I Read Both Codebases TwiceHermes Agent 与 OpenClaw：我把两个代码库都读了两遍

In June I had to pick a runtime for a fleet of agents on a VPS. Hermes Agent or OpenClaw. Most comparisons I found read the README and counted stars, so I cloned both repos and read the source instead. The answer was clean: OpenClaw optimized for reach, Hermes for depth. I picked Hermes and built on it.六月时，我需要为 VPS 上的一组智能体选择运行时，候选者是 Hermes Agent 或 OpenClaw。我发现大多数对比文章只是读了 README 并数了数 Star 数，于是我克隆了两个仓库并亲自阅读了源码。答案很明确：OpenClaw 优化的是覆盖范围，而 Hermes 优化的是深度。最终我选择了 Hermes 并基于它进行开发。

On October 1 I read both codebases again. Three of my June conclusions are now wrong. The gap I based my decision on mostly closed, and what is left is a different, more useful comparison.10 月 1 日，我再次阅读了这两个代码库。我六月得出的结论中，有三个现在已经错了。我当初做决策所依据的差距大部分已经消失，剩下的则是一个完全不同、也更有意义的对比视角。

The short answer简短的回答

Pick OpenClaw if you want a personal assistant that lives on your devices: native apps for macOS, iOS, Android and Linux, a wake word, on-device voice, and a heartbeat that wakes it on its own. Pick Hermes if you want an agent runtime that lives on a server: seven execution backends including serverless sandboxes, a Mixture-of-Agents mode, and a model lab behind it.如果你想要一个生活在你设备上的个人助理，请选择 OpenClaw：它提供 macOS、iOS、Android 和 Linux 的原生应用、唤醒词功能、端侧语音处理，以及能自动唤醒的“心跳”机制。如果你想要一个运行在服务器上的智能体运行时，请选择 Hermes：它拥有包括无服务器沙箱在内的七个执行后端、智能体混合模式（Mixture-of-Agents）以及背后的模型实验室支持。

Everything in between, they now both have. Learning loops, file-based memory, a way for the model to script its own tool calls, recursive subagents, MCP in both directions. Five months ago that sentence was false.至于两者之间的其他功能，现在它们都有了。学习循环、基于文件的记忆、模型编写工具调用脚本的能力、递归子智能体，以及双向 MCP 支持。五个月前，这句话还是错的。

Two kinds of powerful两种“强大”

“Which one is more powerful” is the question everyone asks and the one that cannot be answered as asked. There are two senses of powerful in play. Reach: how many places the agent lives, how many integrations it has, how many people build on it. Depth: how much control you have over where it executes, how it manages context, how it learns.“哪一个更强大”是每个人都会问的问题，但这个问题本身就无法回答。这里存在两种不同层面的“强大”。覆盖范围（Reach）：智能体能驻留多少地方、集成了多少功能、有多少人在其上进行开发。深度（Depth）：你对智能体执行位置的控制力、它如何管理上下文、以及它如何学习。

In June, OpenClaw won reach without a contest and Hermes won depth without a contest. In October, OpenClaw still wins reach. Depth is now mostly a tie, with a few places where Hermes is still ahead.六月时，OpenClaw 在覆盖范围上遥遥领先，Hermes 在深度上毫无争议地胜出。到了十月，OpenClaw 依然在覆盖范围上领先，而深度方面现在基本打平，仅在少数几个地方 Hermes 仍略占优势。

What changed between June and October六月到十月之间发生了什么

My June conclusions, checked again in October我六月的结论，在十月重新核对

What I concluded in June	What the source says in October
OpenClaw does not create skills on its own	Wrong now. Self-learning is on by default: after substantial work, a detached background review proposes a reusable skill, and the Skill Workshop checks it before applying it.
Only Hermes lets the model script its tools	Wrong now. OpenClaw's Code Mode hides the tool schemas and lets the model write JavaScript against the catalog, in a Node vm or a hardened QuickJS guest.
OpenClaw subagents stop at depth one	Wrong now. Recursive delegation to depth five, five children per agent, eight concurrent.
OpenClaw has far more messaging channels	Parity. About 26 in-repo channels each. Hermes even ships WeChat and Yuanbao natively, which OpenClaw leaves to external plugins.
Hermes has six execution backends and 28 model providers	Seven backends, with Vercel Sandbox added, and 38 provider plugins.

The pattern is plain. OpenClaw spent the summer closing the depth gap. Its 2.0 release in August rebuilt the control UI, moved sessions to SQLite, and shipped the Skill Workshop. Hermes spent the same months widening its surface: Bot Mode for named agents in group chats, cron jobs that keep memory between runs, live steering of running subagents, an MCP command center. Each copied the other’s strengths. Hermes’ own MCP server says in its docstring that it matches OpenClaw’s channel bridge surface.模式很明显。OpenClaw 在整个夏天都在弥补深度方面的差距。八月份发布的 2.0 版本重构了控制界面，将对话会话迁移至 SQLite，并推出了“技能工坊”（Skill Workshop）。Hermes 在这几个月里则致力于拓宽其应用面：为群聊中指定智能体提供“机器人模式”、在多次运行间保持记忆的定时任务、对运行中子智能体的实时干预、以及 MCP 指挥中心。双方都在复制对方的优势。Hermes 自己的 MCP 服务器在文档字符串中甚至明确表示，它对标了 OpenClaw 的通道桥接界面。

Hermes Agent vs OpenClaw, side by sideHermes Agent 与 OpenClaw 对比

October 20262026 年 10 月

Both are model-agnostic, both speak MCP as client and server, both run local models.

	Hermes Agent	OpenClaw
Design center	Agent runtime on a server	Personal assistant on your devices
Behind it	Nous Research, a model lab. MIT.	OpenClaw Foundation, a 501(c)(3). MIT.
GitHub	~250k stars, ~48k open issues and PRs	~391k stars, ~9k open issues and PRs
Surfaces	CLI, terminal UI, desktop app, web dashboard, API	Native macOS, iOS, Android and Linux apps, wake word, on-device voice
Wakes on its own	Cron jobs	Heartbeat every 30 minutes by default, plus cron
Where commands run	Local, Docker, SSH, Singularity, Modal, Daytona, Vercel Sandbox	Host by default; opt-in sandbox in Docker, Podman, SSH, OpenShell or Crabbox
Model scripts its tools	Python over RPC, 50 KB of stdout returned	Code Mode, JavaScript against a hidden catalog
Learning	Background review every 10 turns, weekly curator	Self-learning after substantial work, Skill Workshop
Memory	MEMORY.md, USER.md, full-text session search, 7 external providers	MEMORY.md, USER.md, daily notes, dreaming, session search, Honcho
Mixture of agents	Yes, /moa	No
Skills	SKILL.md, agentskills.io format	SKILL.md, ClawHub registry, 164 extensions

Where OpenClaw still leadsOpenClaw 依然领先的地方

Presence. OpenClaw has native apps for four platforms, a wake-word daemon that runs on-device with zero network traffic, on-device text to speech on the Mac, and a canvas panel. Hermes has a desktop app, a terminal UI and a web dashboard, and no native mobile app. If the agent should live on your phone, this is not close. Windows is the gap on OpenClaw’s side: the gateway runs there, the companion app does not exist yet.存在感。OpenClaw 为四个平台提供了原生应用，拥有一个在设备端运行且零网络流量的唤醒词守护进程、Mac 端的端侧文本转语音功能，以及一个画布面板。Hermes 拥有桌面应用、终端 UI 和 Web 仪表盘，但没有原生移动应用。如果智能体需要驻留在你的手机上，那两者根本没法比。OpenClaw 目前的短板在于 Windows：虽然网关可以在 Windows 上运行，但配套应用尚未推出。

Ambient autonomy. OpenClaw’s heartbeat wakes the agent every 30 minutes by default, one hour when you use Anthropic OAuth, and lets it decide whether something needs doing. Hermes only wakes on cron or when something talks to it. For a personal assistant, the heartbeat is the feature. For a fleet driven by an orchestrator, it is a cost you do not need.环境自主性。OpenClaw 的“心跳”机制默认每 30 分钟唤醒一次智能体（使用 Anthropic OAuth 时为一小时），并让它自行决定是否有任务需要处理。Hermes 仅在定时任务触发或有外部交互时才唤醒。对于个人助理而言，心跳是核心功能；但对于由调度器驱动的智能体集群来说，这是一种不必要的开销。

Governance and process. OpenClaw is run by an independent foundation, with OpenAI, NVIDIA, GitHub, Vercel and Red Hat among its sponsors. The README is explicit that OpenAI is a donor, not an owner. Its 2.0 release credited 987 contributors. It carries about a fifth of Hermes’ open issue and pull request count with more stars. That ratio tells you something about how each project absorbs its own growth.治理与流程。OpenClaw 由一个独立基金会运营，赞助商包括 OpenAI、NVIDIA、GitHub、Vercel 和 Red Hat。README 中明确指出 OpenAI 是捐赠者而非所有者。其 2.0 版本致谢了 987 名贡献者。在 Star 数更多的情况下，它的开放 Issue 和 Pull Request 数量仅为 Hermes 的五分之一左右。这个比例说明了这两个项目在消化自身增长方面的不同能力。

Ecosystem. 164 extensions, 49 bundled skills, and ClawHub for installing and publishing more.生态系统。拥有 164 个扩展、49 个捆绑技能，以及用于安装和发布更多内容的 ClawHub。

Where Hermes still leadsHermes 依然领先的地方

Where your commands run. Hermes can execute in seven backends, three of them cloud sandboxes: Modal, Daytona and Vercel Sandbox. OpenClaw runs tools on the host unless you turn sandboxing on, and its sandbox options are Docker, Podman, SSH, OpenShell and Crabbox. If you want the agent on one machine and its commands somewhere disposable, Hermes gives you more places to put them.命令执行环境。Hermes 可以在七个后端执行任务，其中三个是云端沙箱：Modal、Daytona 和 Vercel Sandbox。OpenClaw 除非开启沙箱，否则默认在宿主机上运行工具，其沙箱选项包括 Docker、Podman、SSH、OpenShell 和 Crabbox。如果你希望智能体运行在一台机器上，而命令在临时的隔离环境中执行，Hermes 提供了更多选择。

Mixture of agents. Hermes can send one question to several models and aggregate the answers. I found no equivalent in OpenClaw.智能体混合模式。Hermes 可以将一个问题发送给多个模型并汇总答案，我在 OpenClaw 中没找到对应的功能。

A model lab behind it. Hermes ships a batch runner and a trajectory compressor that write tool-calling trajectories in Nous’ own format. The runtime doubles as a way to train the lab’s models. That gives the project a reason to exist beyond its users, and it shows in how deep the loop primitives go.背后的模型实验室。Hermes 附带了一个批处理运行器和一个轨迹压缩器，它们以 Nous 的自有格式编写工具调用轨迹。该运行时同时也作为实验室模型训练的一种手段。这赋予了该项目在用户需求之外存在的理由，这一点也体现在其循环原语的深度上。

Running subagents you can steer. Both delegate recursively. Hermes runs up to 10 children at once by default and, since v0.21, lets you list, steer and stop them while they run.可干预的子智能体运行。两者都支持递归委托。Hermes 默认同时运行最多 10 个子任务，并且自 v0.21 版本起，允许你在它们运行时进行列表查看、干预和停止。

What each one costs to run各自的运行成本

Neither is cheap by default. Both can be made cheap. In both cases the cost is a setting, not a property of the project.默认情况下两者都不便宜，但都可以优化。在这两种情况下，成本都是配置问题，而非项目本身的属性。

OpenClaw’s bill comes from the heartbeat. Each beat is a full agent turn. The docs put it plainly: with isolatedSession on, a heartbeat drops from about 100k tokens to about 2k to 5k per run, because it stops resending the conversation history. Turn on lightContext too and it skips the workspace bootstrap files.OpenClaw 的账单主要来自心跳机制。每一次心跳都是一次完整的智能体轮次。文档对此说明得很清楚：开启 isolatedSession 后，单次心跳的消耗会从约 10 万 Token 降至约 2000 到 5000 Token，因为它停止了重复发送对话历史。同时开启 lightContext 还可以跳过工作空间引导文件的加载。

Hermes’ bill comes from the learning loop. With the same model, the background review replays the whole session history and leans on the prompt cache to keep that affordable. Route the review to a different model and it collapses the history to a digest. An issue arguing the full-history review is too expensive has been open since June with no reply.Hermes 的账单则来自学习循环。使用相同模型时，后台回顾会重放整个会话历史，并依赖提示词缓存（Prompt Cache）来保持成本可控。如果将回顾任务路由给另一个模型，它会将历史压缩为摘要。关于“全量历史回顾成本过高”的 Issue 自六月以来一直处于开放状态，且没有回复。

The only number that matters is yours. Run the same task through both, on the same model, and read the token counts. I learned that the hard way when a “light” mode in my own setup cost six times more than the full one. The token cost guide has the rest of the method.唯一重要的数字是你自己的。在同一个模型上运行相同的任务，并查看 Token 统计。我曾吃过亏，当时我配置中的“轻量”模式成本竟然是全量模式的六倍。Token 成本指南中包含了其余的方法论。

Both are still young software两者都还是年轻的软件

Read the open issues before you put either one on a box you depend on. On OpenClaw: a gateway whose memory grows from 350 MB to 15.5 GB over days until it is OOM-killed (#91588), a SQLite write-ahead log that grows to gigabytes and blocks startup (#143524), an MCP server init timeout that crashes the gateway (#144911), and a September release that turned one user’s stable setup into an eight-hour recovery (#153257).在你将它们部署到关键生产环境之前，请务必阅读开放的 Issue。关于 OpenClaw：网关内存会在几天内从 350MB 涨到 15.5GB 直至触发 OOM 被杀（#91588）、SQLite 预写日志增长到数 GB 并阻塞启动（#143524）、MCP 服务器初始化超时导致网关崩溃（#144911），以及九月份的版本更新导致某位用户的稳定环境耗费了八小时才恢复（#153257）。

On Hermes: v0.21.0 shipped session-database regressions that took six pull requests and 44 closed issues to settle, a gateway executor with no timeout freezes the gateway for 120 seconds (#101033), and the dashboard leaks memory until clients get OOM-killed (#80527).关于 Hermes：v0.21.0 引入了会话数据库回归问题，耗费了 6 个 PR 和 44 个关闭的 Issue 才解决；网关执行器在没有超时设置的情况下会导致网关冻结 120 秒（#101033）；仪表盘存在内存泄漏，导致客户端被 OOM 杀掉（#80527）。

Same advice for both. Pin a version. Upgrade one box first. Keep the logs somewhere you can read them without asking the agent.给两者的建议相同：锁定版本。先升级一台机器。把日志存放在你可以直接查看的地方，不要依赖智能体去查。

The claims the code killed被代码打破的那些断言

The method matters more than the verdict, because the verdict expires. In June three marketing claims fell apart once I opened the source. “3,200 MCP tools” turned out to be the size of the MCP ecosystem, not anything in either repo. The star gap described popularity, not what each runtime can do. And “OpenClaw burns more tokens” turned out to be a heartbeat setting. I wrote that up in the Lab as the marketing numbers were the story someone wanted me to tell.方法论比结论更重要，因为结论是有时效的。六月时，当我打开源码，三个营销宣传点就不攻自破了。“3200 个 MCP 工具”实际上是整个 MCP 生态系统的规模，而非这两个仓库中的任何一个。Star 数的差距反映的是知名度，而非运行时本身的功能。而“OpenClaw 消耗更多 Token”实际上是一个心跳设置问题。我当时在实验室对此进行了记录，因为那些营销数字正是别人希望我传播的故事。

In October the code killed three of my own conclusions. That is the same lesson pointed back at me. A comparison read from the source is only true for the commit you read.十月时，代码打破了我自己的三个结论。这对我来说是同样的教训。基于源码的对比，其有效性仅限于你阅读的那个 Commit。

Why I still run Hermes为什么我依然使用 Hermes

My case is specific: a VPS, a cost-sensitive budget, and a fleet of work agents that an orchestrator wakes on its own schedule. I do not need a heartbeat, because the orchestrator is the heartbeat. I want commands to execute somewhere disposable. And my adapter, my agents’ instruction bundles and my MCP pooler are all built around Hermes. For that case, Hermes is still the right call.我的情况很特殊：一台 VPS、严格的预算限制，以及一组由调度器按计划唤醒的工作智能体。我不需要心跳，因为调度器本身就是心跳。我希望命令在临时的环境中执行。而且我的适配器、智能体的指令包以及 MCP 池化器都是围绕 Hermes 构建的。对于这种情况，Hermes 依然是正确的选择。

If I were starting from zero today, the decision would be closer than it was in June. And it does not have to be one or the other. They plug into the same orchestrators, so you can route by role: Hermes for the work that runs on a server, OpenClaw for the assistant that lives on your phone.如果我今天从零开始，决策过程会比六月时更艰难。而且这并非非此即彼的选择。它们可以接入相同的调度器，因此你可以根据角色进行路由：用 Hermes 处理服务器端的工作任务，用 OpenClaw 处理驻留在手机上的个人助理。

Looking for an OpenClaw alternative?在寻找 OpenClaw 的替代品？

“OpenClaw alternative” alone gets about 1,900 searches a month in the US. Before you switch, check that switching fixes your actual reason.在美国，“OpenClaw 替代品”每月大约有 1900 次搜索。在切换之前，请确认切换是否真的能解决你的根本痛点。

Before you replace OpenClaw在替换 OpenClaw 之前

• ✓Required:
  The heartbeat bill is too high.Turn on isolatedSession and lightContext first. The docs put a tuned heartbeat at 2k to 5k tokens instead of about 100k.
  ✓必要操作：心跳账单太高。请先开启 isolatedSession 和 lightContext。文档显示，优化后的心跳单次消耗在 2k 到 5k Token，而不是约 10 万。
• ✓Required:
  You do not want an agent running commands on your host.Turn sandboxing on, or move to Hermes and execute in Docker, SSH or a serverless sandbox. Either way, give it its own machine and keys you can rotate.
  ✓必要操作：你不希望智能体在宿主机上运行命令。请开启沙箱，或迁移到 Hermes 并在 Docker、SSH 或无服务器沙箱中执行。无论如何，请给它分配独立的机器和可轮换的密钥。
• ✓Required:
  The gateway keeps running out of memory.It is a known open bug. Pin a version that behaves. Hermes has its own memory leak in the dashboard, so moving does not make the class of problem go away.
  ✓必要操作：网关内存持续耗尽。这是一个已知的开放 Bug。请锁定一个表现稳定的版本。Hermes 的仪表盘也有内存泄漏，所以迁移并不能解决这类问题。
• ✓Required:
  You want it on a server, driven by an orchestrator or cron.This is the honest reason to pick Hermes. Its design center is the agent that runs when nobody is talking to it.
  ✓必要操作：你希望它运行在服务器上，由调度器或定时任务驱动。这是选择 Hermes 的诚实理由。它的设计核心就是那种在无人交互时也能运行的智能体。
• ✓Required:
  What you actually want is a coding agent.Neither is the best tool for that. Use Claude Code, OpenCode or pi inside the repo.
  ✓必要操作：你真正想要的是一个编程智能体。两者都不是最佳工具。请在仓库内使用 Claude Code、OpenCode 或 pi。

How to choose如何选择

Pick OpenClaw if如果满足以下条件，请选择 OpenClaw：

1. 01The agent should live on your phone and laptop01 智能体需要驻留在你的手机和笔记本电脑上
2. 02You want voice, a wake word and native apps02 你需要语音、唤醒词和原生应用
3. 03It should wake on its own and decide what to do03 它需要自动唤醒并自行决定做什么
4. 04You value a foundation and a large ecosystem04 你看重基金会背景和庞大的生态系统
5. 05You are not on Windows, or only need the gateway there05 你不在 Windows 上运行，或者仅需在 Windows 上运行网关

Pick Hermes if如果满足以下条件，请选择 Hermes：

1. 01The agent should live on a server01 智能体需要驻留在服务器上
2. 02Commands should run in disposable sandboxes02 命令需要在临时沙箱中执行
3. 03An orchestrator or cron decides when it wakes03 由调度器或定时任务决定唤醒时间
4. 04You want Mixture-of-Agents or live subagent steering04 你需要智能体混合模式或对子智能体的实时干预
5. 05You will read Python and logs when it breaks05 出问题时你愿意阅读 Python 代码和日志

Hermes Agent vs OpenClaw, quick answersHermes Agent 与 OpenClaw，快速问答

Is Hermes Agent better than OpenClaw?+

Not in general. They have different design centers. OpenClaw is better as a personal assistant on your devices. Hermes is better as a runtime on a server, especially when commands should run in disposable sandboxes.并非一概而论。它们有不同的设计中心。OpenClaw 更适合作为你设备上的个人助理。Hermes 更适合作为服务器上的运行时，特别是在命令需要在临时沙箱中执行时。

Since August most core capabilities overlap: both learn skills, keep file-based memory, let the model script its tools and delegate recursively.自八月以来，大多数核心功能已经重叠：两者都能学习技能、保持基于文件的记忆、让模型编写工具脚本并递归委托任务。

Does OpenClaw learn skills like Hermes?+

Yes, since its 2.0 release in August 2026. Self-learning is on by default: after substantial work, a background review proposes a reusable skill, and the Skill Workshop checks proposals before applying them. Hermes runs its review every 10 turns or tool iterations and archives unused skills weekly.是的，自 2026 年 8 月发布的 2.0 版本起。自我学习是默认开启的：在完成大量工作后，后台回顾会提出一个可重用的技能，而“技能工坊”会在应用前检查这些建议。Hermes 每 10 次轮次或工具迭代运行一次回顾，并每周归档未使用的技能。

Which one uses fewer tokens?+

It depends on configuration, not on the project. OpenClaw's heartbeat is the main cost and drops from about 100k to about 2k to 5k tokens per run with isolatedSession. Hermes' main extra cost is the background review replaying session history.这取决于配置，而非项目本身。OpenClaw 的心跳是主要成本，通过 isolatedSession 可以从约 10 万 Token 降至约 2000 到 5000 Token。Hermes 的主要额外成本是重放会话历史的后台回顾。

Measure both on the same task and the same model before you decide.在决定之前，请在相同的任务和模型上对两者进行测量。

Can I run Hermes and OpenClaw together?+

Yes. They are not mutually exclusive, and orchestrators like Paperclip have adapters for both. A sensible split is Hermes for server-side work agents and OpenClaw for the personal assistant on your devices.是的。它们并不互斥，像 Paperclip 这样的调度器同时适配两者。一个合理的分配是：Hermes 用于服务器端的工作智能体，OpenClaw 用于设备上的个人助理。

Is OpenClaw owned by OpenAI?+

No. OpenClaw is MIT-licensed and run by the OpenClaw Foundation, an independent 501(c)(3). OpenAI is one of its sponsors, and the README states it is a donor, not an owner.不是。OpenClaw 使用 MIT 协议，由独立的 501(c)(3) 非营利机构 OpenClaw 基金会运营。OpenAI 是其赞助商之一，README 中明确说明它是捐赠者而非所有者。