Nous research raises $75m at $1.5b valuation for hermes agent Nous Research, creator of the open-source Hermes AI agent, is closing a new funding round led by Robot Ventures at a $1.5 billion valuation, sources say....
seen from China

seen from United States
seen from United States
seen from United States

seen from Malaysia
seen from United States
seen from United States

seen from United States
seen from United States
seen from United States
seen from China

seen from United Kingdom
seen from United States
seen from China

seen from Canada
seen from China

seen from United States
seen from Italy
seen from United States

seen from Poland
Nous research raises $75m at $1.5b valuation for hermes agent Nous Research, creator of the open-source Hermes AI agent, is closing a new funding round led by Robot Ventures at a $1.5 billion valuation, sources say....

Anya is live and ready to show you everything. Watch her strip, dance, and perform exclusive shows just for you. Interact in real-time and make your fantasies come true.
Free to watch • No registration required • HD streaming
Hermes Agent 模型切换指南:Claude Opus 4.8、GPT-5.5、DeepSeek V4 Flash 怎么选?
Hermes Agent 模型切换指南:Claude Opus 4.8、GPT-5.5、DeepSeek V4 Flash 怎么选?
Hermes Agent 接入 Crazyrouter 后,一个很实用的能力就是:你可以在同一个工具里切换不同模型。
比如:
/model claude-opus-4-8 /model gpt-5.5 /model deepseek-v4-flash
这篇文章讲清楚:不同模型适合什么任务,以及如何通过 crazyrouter-hermes 一键配置 Hermes Agent。
脚本地址:
https://github.com/xujfcn/crazyrouter-hermes
默认接口:
https://cn.crazyrouter.com/v1
---
为什么要在 Hermes Agent 里配置多个模型?
AI Agent 的任务并不都一样。
有些任务需要强推理,有些任务需要快速回答,有些任务只是简单解释代码。
如果所有任务都用同一个模型,要么太贵,要么太慢,要么能力不够。
更合理的方式是:
| 场景 | 推荐模型 |
|---|---|
| 复杂代码分析 | claude-opus-4-8 |
| 通用开发问答 | gpt-5.5 |
| 快速轻量任务 | deepseek-v4-flash |
| 多模态/综合任务 | gemini-3.1-pro |
| 兼容性测试 | gpt-4o |
---
一键配置 Hermes Agent
Linux / macOS / WSL2:
curl -fsSL https://raw.githubusercontent.com/xujfcn/crazyrouter-hermes/main/setup.sh | bash
Windows PowerShell:
irm https://raw.githubusercontent.com/xujfcn/crazyrouter-hermes/main/setup.ps1 | iex
脚本会默认写入:
OPENAI_BASE_URL=https://cn.crazyrouter.com/v1
并生成 Hermes 配置:
model: provider: "custom" default: "claude-opus-4-8" base_url: "https://cn.crazyrouter.com/v1"
---
Claude Opus 4.8:适合复杂任务
适合用在:
多文件代码理解;
复杂 bug 分析;
长任务规划;
架构设计;
Agent 多步骤任务。
切换命令:
/model claude-opus-4-8
示例提示词:
帮我分析这个项目的目录结构,并给出重构建议
如果你需要质量优先,先用它。
---
GPT-5.5:适合通用开发任务
适合用在:
代码解释;
文档生成;
快速问答;
单文件脚本修改;
API 示例生成。
切换命令:
/model gpt-5.5
示例提示词:
把这个 Python 脚本改成支持命令行参数
如果你不确定用哪个模型,可以先用 GPT-5.5。
---
DeepSeek V4 Flash:适合快速轻量任务
适合用在:
简单代码解释;
快速摘要;
日常命令行问题;
轻量问答;
低复杂度脚本生成。
切换命令:
/model deepseek-v4-flash
示例提示词:
解释一下这个报错是什么意思,并给出两种可能原因
它适合“先快速问一下”的场景。
---
Gemini 3.1 Pro:适合综合任务
切换命令:
/model gemini-3.1-pro
适合用在:
长上下文理解;
多类型信息整理;
综合分析;
需要换模型对比答案的任务。
---
GPT-4o:适合兼容性测试
切换命令:
/model gpt-4o
阅读完整教程
Hermes Agent 配置 Base URL 指南:为什么必须写 https://cn.crazyrouter.com/v1
Hermes Agent 配置 Base URL 指南:为什么必须写 https://cn.crazyrouter.com/v1
配置 AI Agent 工具时,Base URL 是最容易被忽略、但最容易导致失败的参数。
很多人有 API Key,也知道模型名,但 Hermes Agent 还是请求失败。最后发现问题只有一个:Base URL 写错了。
这篇文章专门讲 Hermes Agent 的 Base URL 配置,尤其是接入 Crazyrouter 时为什么推荐使用:
https://cn.crazyrouter.com/v1
配套脚本:
https://github.com/xujfcn/crazyrouter-hermes
---
Base URL 是什么?
Base URL 是客户端发送请求时使用的 API 根地址。
比如:
https://cn.crazyrouter.com/v1
当 Hermes Agent 调用聊天接口时,最终请求路径通常会变成:
https://cn.crazyrouter.com/v1/chat/completions
如果你少写 /v1,就可能变成:
https://cn.crazyrouter.com/chat/completions
这就很容易 404。
---
为什么要用 cn.crazyrouter.com?
对国内开发者来说,推荐默认使用:
https://cn.crazyrouter.com/v1
原因很简单:
1. 接口路径更适合国内访问;
2. Hermes Agent 使用 OpenAI-compatible API;
3. /v1 是很多 SDK 和工具默认期望的 API 版本路径;
4. 配置统一后,模型切换更简单。
---
手动配置方式
编辑:
~/.hermes/.env
写入:
OPENAI_API_KEY=sk-your-crazyrouter-key OPENAI_BASE_URL=https://cn.crazyrouter.com/v1
再编辑:
~/.hermes/config.yaml
写入:
model: provider: "custom" default: "claude-opus-4-8" base_url: "https://cn.crazyrouter.com/v1"
如果你不想手动改,可以直接用脚本。
---
一键配置脚本
Linux / macOS / WSL2:
curl -fsSL https://raw.githubusercontent.com/xujfcn/crazyrouter-hermes/main/setup.sh | bash
Windows PowerShell:
irm https://raw.githubusercontent.com/xujfcn/crazyrouter-hermes/main/setup.ps1 | iex
Windows CMD:
curl -o setup.bat https://raw.githubusercontent.com/xujfcn/crazyrouter-hermes/main/setup.bat setup.bat
脚本会自动把 Base URL 写成:
https://cn.crazyrouter.com/v1
---
常见错误 1:少了 /v1
错误:
https://cn.crazyrouter.com
正确:
https://cn.crazyrouter.com/v1
少了 /v1 时,请求路径会不完整。
---
常见错误 2:重复写了 /v1
错误:
https://cn.crazyrouter.com/v1/v1
正确:
https://cn.crazyrouter.com/v1
重复 /v1 也会导致接口路径错误。
---
常见错误 3:模型名和接口不匹配
Base URL 正确后,还要确认模型名可用。
推荐先测试:
claude-opus-4-8 gpt-5.5 deepseek-v4-flash
进入 Hermes 后切换:
/model claude-opus-4-8 /model gpt-5.5 /model deepseek-v4-flash
阅读完整教程
Hermes Agent 一键接入 Crazyrouter:用国内接口跑 Claude Opus 4.8、GPT-5.5 和 DeepSeek V4 Flash
Hermes Agent 一键接入 Crazyrouter:用国内接口跑 Claude Opus 4.8、GPT-5.5 和 DeepSeek V4 Flash
如果你正在用 Hermes Agent 做本地 AI 助手,最容易卡住的地方通常不是模型能力,而是配置:API Key 写在哪里?Base URL 要不要带 /v1?默认模型怎么选?配置错了怎么恢复?
crazyrouter-hermes 这个脚本就是为了解决这类问题:一条命令把 Hermes Agent 接到 Crazyrouter,并默认使用国内接口。
GitHub 脚本:
https://github.com/xujfcn/crazyrouter-hermes
默认接入地址:
https://cn.crazyrouter.com/v1
默认推荐模型:
claude-opus-4-8 gpt-5.5 deepseek-v4-flash
---
为什么 Hermes Agent 需要自定义模型配置?
Hermes Agent 是一个可以在本地运行的 AI Agent 工具。它能帮你处理代码、命令行、文件和日常开发任务。
但要让 Hermes Agent 使用你想要的模型,通常需要配置:
| 配置项 | 作用 |
|---|---|
| OPENAI_API_KEY | API 鉴权 |
| OPENAI_BASE_URL | OpenAI-compatible API 地址 |
| model.default | 默认模型 |
| model.provider | 模型供应商类型 |
手动配置当然可以,但新手很容易把 base_url 写错。比如少了 /v1:
https://cn.crazyrouter.com
正确应该是:
https://cn.crazyrouter.com/v1
一键脚本的价值就是把这些容易出错的步骤固定下来。
---
Linux / macOS / WSL2 一键安装
执行:
curl -fsSL https://raw.githubusercontent.com/xujfcn/crazyrouter-hermes/main/setup.sh | bash
脚本会做这些事:
1. 检查 Hermes Agent 是否存在;
2. 创建 ~/.hermes 目录;
3. 提示输入 Crazyrouter API Key;
4. 选择默认模型;
5. 写入 .env;
6. 写入或更新 config.yaml;
7. 可选测试连接。
执行完成后,启动 Hermes:
hermes
然后可以在 Hermes 内切换模型:
/model claude-opus-4-8 /model gpt-5.5 /model deepseek-v4-flash
---
Windows PowerShell 一键安装
Windows 用户可以执行:
irm https://raw.githubusercontent.com/xujfcn/crazyrouter-hermes/main/setup.ps1 | iex
如果你担心远程执行脚本,可以先下载:
iwr https://raw.githubusercontent.com/xujfcn/crazyrouter-hermes/main/setup.ps1 -OutFile setup.ps1
打开看一遍源码,再执行:
.\setup.ps1
这更适合公司电脑或对安全比较敏感的环境。
---
脚本写入了哪些文件?
主要是两个文件:
~/.hermes/.env ~/.hermes/config.yaml
.env 示例:
OPENAI_API_KEY=sk-your-crazyrouter-key OPENAI_BASE_URL=https://cn.crazyrouter.com/v1
config.yaml 示例:
model: provider: "custom" default: "claude-opus-4-8" base_url: "https://cn.crazyrouter.com/v1"
脚本覆盖旧配置前会生成 .bak 备份。
如果配置错了,可以用备份恢复:
~/.hermes/.env.bak ~/.hermes/config.yaml.bak
---
推荐模型怎么选?
1. Claude Opus 4.8
适合:
复杂代码理解;
长任务规划;
多步骤 Agent 工作流;
深度分析。
切换命令:
/model claude-opus-4-8
2. GPT-5.5
适合:
通用开发任务;
快速问答;
代码补全;
文档生成。
切换命令:
阅读完整教程
create your own hermes agents

Anya is live and ready to show you everything. Watch her strip, dance, and perform exclusive shows just for you. Interact in real-time and make your fantasies come true.
Free to watch • No registration required • HD streaming
This AI Agent Learns on Its Own… Here’s THE Hermes AI Agent
Hermes Agent, launched February 25, 2026 by Nous Research, is the fastest-growing open-source AI agent of 2026 — hitting 110,000 GitHub stars in ten weeks. Unlike OpenClaw, which executes tasks and stays static, Hermes runs a closed self-learning loop that makes the agent measurably better after every session. In this guide, we’ll cover what Hermes Agent is, how its self-learning loop works, how to set it up, and where it outperforms OpenClaw in real workflows.
What Makes Hermes Agent Different
Hermes Agent is a self-hosted, model-agnostic AI agent built around one principle: the agent improves through use, not configuration.
It was built by Nous Research — an open-source-first collective known for the Hermes model series, Atropos RL environments, and DisTrO distributed training. Their previous work on reasoning-capable, tool-using models fed directly into the agent’s architecture.
Where most agents (including OpenClaw) rely on human-written skill files and static memory, Hermes uses a closed feedback loop. Every task the agent completes becomes training data for its own future performance.
📌 Core design principle: Do → Learn → Improve. Repeat every session.
How the Self-Learning Loop Works
This is the architectural bet that separates Hermes from everything else in 2026.
After completing a task, the agent runs an internal review cycle:
Experience: Agent completes a multi-step workflow
Extraction: It identifies reusable patterns from what it executed
Skill Creation: Writes a new Markdown skill file automatically
Refinement: That skill self-improves every time it’s used
Nudge: Every 15 tasks, the agent reviews and updates its own knowledge base
The result: an agent handling customer queries in March will perform measurably better by June — with no manual updates. Benchmarks from early 2026 show Hermes recovered from task errors 22% more effectively than OpenClaw in Long-Horizon Task tests.
Skills follow the open agentskills.io standard — not a proprietary format, keeping you portable.
How to Use Hermes Agent: Full Set-up
Hermes Agent is written in Python and runs on your own machine or server. Here’s the full Hermes Agent installation to get started:
Prerequisites
Python 3.11+
Git
API key for your chosen LLM provider (Anthropic, OpenAI, etc.)
Telegram bot token (or other messaging interface)
Installation Steps
Step 1: Clone the repositorygit clone https://github.com/NousResearch/hermes-agent cd hermes-agent
Read more at How to Use Hermes Agent? OpenClaw with Self-Learning Loop
AI agent 2026 đã thay đổi luật chơi: đừng chọn sai công cụ
Không biết 3 AI Agent này năm 2026 thì đừng gọi mình là dân tech Kim Jongwook · 2026-04-09 TL;DR Ba AI agent dẫn đầu 2026: OpenClo, Hermes Agent, Claude Code — mỗi bên thống trị một mảng riêng. Hermes Agent thắng OpenClo ở khả năng tự học và ghi nhớ lâu dài, tạo chênh lệch năng suất lớn theo thời gian. Claude Code đứng số 1 SWE-bench với 80,9% và first-pass accuracy 95%, gần như độc chiếm mảng…
AI Agents 2026: Hermes vs OpenClo vs Claude Code Exposed
If You Don’t Know These 3 AI Agents, You’re Already Behind Kim Jongwook · 2026-04-09 TL;DR OpenClo, Hermes Agent, and Claude Code occupy three distinct niches in the 2026 AI agent landscape. Hermes Agent’s self-evolving learning loop and database memory beat OpenClo for long-term productivity. Claude Code leads coding agents with 80.9% SWE-bench and 95% first-pass accuracy in production…