Author｜Jacob Zhao @ IOSG

The phenomenal growth of Hermes is not due to exclusive technologies that cannot be replicated by the OpenClaw principle, but rather because it precisely closed a "challenger growth system" during the critical window when the personal Agent category was forming: leveraging the mature user pool educated by OpenClaw and establishing "Delegation Trust," which offers a more authentic experiential difference than the "self-evolution" narrative. As professional execution Agents become increasingly powerful, users still need a long-term online and trustworthy housekeeper.

Opening the public application leaderboard of OpenRouter, Hermes Agent ranks first across the entire platform with a token usage of 30.5 trillion, while also ranking first in the categories of Productivity, Coding Agents, Personal Agents, and CLI Agents, significantly outpacing well-known Agents like OpenClaw and Claude Code.

▲ Figure 1 · Historical data snapshot of Hermes Agent on OpenRouter (captured on August 4, 2026; dynamic page data will change over time)

Although OpenRouter's statistical criteria cannot cover full industry token consumption from directly connecting to official APIs (such as native subscriptions for Claude or Codex), as the largest AI large model routing and aggregation platform globally, its leaderboard carries significant "weather vane" meaning. While a substantial number of users' core business workflows—complex code generation, architectural design, high-value data analysis—still flow to Claude Code and ChatGPT at the high-end professional task level, Hermes maintains advantages in use cases such as backend automation, message entrance response, long-term online listening, and lightweight task scheduling. As an Agent product created by a Web3 team, Hermes has achieved success in dissemination, community engagement, and usage intensity far exceeding expectations, prompting us to consider:

• Why has Hermes been able to achieve a rebound in inference calls on OpenRouter?
• What are the real distinctions between it and OpenClaw?
• In its relationship with Claude Code and Codex, how does Hermes maintain "differentiated coexistence" rather than "direct competition"?

From Development Framework to Personal AI System—The Path of OpenClaw

Why Early Agent Frameworks Did Not Produce Consumer Products

Before the emergence of OpenClaw, while the Agent domain already had a mature infrastructure, it had fundamental limitations: its adopting unit was "development project enterprise workflows," rather than "individual users." Early frameworks shared common characteristics aimed at developers, producing code or configurations—they built the infrastructure of Agents but did not deliver the Agents themselves. The high engineering threshold led to being perpetually trapped in the "developer tools" phase, lacking a productized closed loop to transform technology into "personal exclusive assets," leaving the "personal Agent product layer" almost entirely blank.

▲ Figure 1 · Six-layer structure of the Agent technology stack (model layer → protocol layer → SDK development framework layer → orchestration runtime layer → execution infrastructure layer → deployment governance layer)

▲ Figure 1 · Historical data snapshot of Hermes Agent on OpenRouter (captured on August 4, 2026; dynamic page data will change over time)

What Did OpenClaw Truly Change?

OpenClaw did not reinvent the Agent Loop or task scheduling technology at its core; its main contribution is a systematic encapsulation at the product level. LangChain addresses "how to build an Agent," while OpenClaw addresses "how to own an Agent." It skips the intermediate layers of the tech stack, integrating scattered framework capabilities into a complete product that individuals can directly configure and use long-term, achieving a fundamental shift in adopting units from "development projects" to "individuals," specifically reflected in six dimensions of product innovation:

• Identity personalization: Endowing Agents with persistent names and identities to break the tool-like feel of stateless API calls.
• Entrance normalization: Using high-frequency communication software like Telegram/WhatsApp as the interaction interface, replacing complex command lines or IDEs.
• Status permanence: Operating as a backend process running online for a long time, achieving a leap from passive "waiting" to active "presence."
• Authority embodiment: Deeply incorporating users' file systems, browsers, terminals, and real-world action capabilities into the operational boundaries of the Agent.
• Capability scalability: Through Skills, Memory, and community plugins, processes are crystallized into reusable capabilities, expanding action boundaries.
• Mental ownership: The most fundamental transformation—users shift from "using an AI tool" to "owning a dedicated digital partner."

Why the Lobster Boom Did Not Form a Second Mentality

The popularity of OpenClaw has spawned numerous imitators. These products address real user issues: cumbersome installation processes, difficult environment configurations, missing channels like WeChat and Feishu, compatibility of domestic models, rapid deployment of cloud hosts, enterprise permission management, automatic updates, and security isolation, among others. They all have their own user bases and reasonable business logic. Yet almost none formed an independent brand mentality—this is because they address the question of "how to use OpenClaw more easily," rather than "where should personal Agents evolve after OpenClaw." The narrative challenger position is extremely scarce in the entire personal Agent market.

Why Ultimately Hermes Outperformed

The Model, Community, and Crypto-native Background of Nous

Nous Research originated from the Discord open-source AI research community in 2022 and officially completed its company operations in 2023. The core founding team includes Jeffrey Quesnelle, Karan Malhotra, Teknium, and Shivani Mitra, covering the following business areas:

• Hermes model series: The most representative open-source model brand of Nous, focusing long-term on model post-training, instruction fine-tuning, and Agent capabilities, establishing a large developer adoption base on Hugging Face.
• DisTrO (Distributed Training Over-the-Internet): Significantly reduces the cross-node communication overhead of distributed training, allowing heterogeneous hardware across regions to participate in collaborative training under internet bandwidth conditions more feasibly.
• Psyche decentralized training network: Further networks DisTrO, coordinating global distributed computing nodes through Solana, allowing GPUs in different networks and hardware environments to jointly participate in large model training.
• Hermes Agent: A personal Agent product launched by Nous for end-users, integrating Hermes models, tool calls, Memory, Skills, messaging channels, and long-term operational capabilities into a resident Agent.

In April 2025, Nous Research completed a $50 million Series A round financing led by Paradigm, with a post-investment token valuation of $1 billion. Before this round of financing, the company had already completed approximately $20 million in early financing from well-known institutions such as Distributed Global, North Island Ventures, and Delphi Digital.

Nous has built a technical closed loop of “Hermes (model capability), DisTrO (distributed training), Psyche (decentralized computing network), and Hermes Agent (personal terminal product).” The release of Hermes Agent is not a temporary fork chasing trends but a strategic extension initiated by Nous to the demand side (real users, tasks, workflows) after long-term accumulation on the supply side (data, models, training, open weights)—this provides a deeper starting point for establishing differentiation compared to ordinary imitations.

OpenClaw's Operation and Maintenance Pain Points Transformed Into Hermes's Growth Engine

Hermes and OpenClaw are not significantly different at the underlying encapsulation (model + tools + Memory + scheduling). Its phenomenal explosion does not rely on technological advantages but precisely closes a systematic growth causal chain: directly inheriting the user pool, which OpenClaw has educated and tortured by operational pain points, through seamless migration of tools, forming the early core growth engine.

Productive Leap: Establishing "Delegation Trust"

The core product assumption of Hermes is addressing "passing operational responsibilities," committing to "absorbing and fixing errors internally within the system":

• Reliability trust: Ensuring task continuity and failure recovery (persistent Kanban, /goal model, tools self-healing).
• Security trust: Preventing overreach, accidental deletions, or data leaks (Approvals flow, sandbox, strict permission boundary).
• Verifiable trust: Proving tasks were truly completed (Completion Contract and Grounded Citations).

Conceptual Analysis: "Self-evolution" (narrative advantage) vs "Autonomous Recovery" (experiential difference)

In Hermes's product narrative, there is a significant difference in product value between "self-evolution" and "autonomous recovery":

• Self-improvement: Essentially a process adaptation based on Memory and Skills. Given that competitors have similar infrastructures, its differentiation lies more in being the first to integrate into a default system with lifecycle management, occupying a narrative advantage of being "growing," rather than having proven and insurmountable technological barriers.
• Autonomous Recovery: This is the current experience difference most worth validating. Thanks to structured error returns and Provider automatic fallback, Hermes can digest failures internally within the system. This system-level stability of "not frequently disturbing users" represents a more direct and perceptible difference in product capability.

Architectural Dividend: Delegation and Supervision Capabilities for Professional Agents

The core value of Hermes lies not in personally executing all professional tasks, but in serving as a control layer (Orchestrator) to fulfill demand completion, decompose tasks, monitor routing, and perform final acceptance. By delegating specific tasks to external CLIs like Claude Code/Codex through built-in Skill, the community has crystallized the practical paradigm of "Hermes control + external CLI as Workers" (e.g., /goal mechanism and oh-my-hermes collaborative tools), reflecting its architectural advantage in raising the complexity ceiling of tasks through scheduling professional Agents.

From Crypto-native to Crypto-invisible: Hermes's Web3 Backend Operating System

Simply attributing Hermes's success to "Web3 background" is overly simplistic. Web3 provides Nous with a set of "organizational operating systems" that are difficult for other AI startup teams to simultaneously obtain, allowing it to enter the mainstream market with a smooth experience of standard AI products:

• Patience of venture capital: Crypto-native capital supports long-term, highly uncertain, and parallel investment routes, enabling Nous to layout models, training, Runtime, and Cloud simultaneously without prematurely converging on a single revenue validation.
• Ready-made user market: Provides a user base familiar with Telegram, servers, APIs, and self-hosting Crypto AI, significantly reducing the cold start education cost and fostering high usage intensity, tutorial dissemination, and Skills contributions.
• User sovereignty values: Adhering to self-hosting, openness, portability, and anti-platform lock-in orientations, directly translating into a MIT License, multi-Provider support, BYOK, and Memory/Skills portable underlying architecture.
• Community R&D and verticalization: Relying on global remote collaboration and open-source culture, users spontaneously become Contributors, Skill authors, and product designers in vertical scenarios.

Hermes exposes almost none of the Crypto to the user front. Using its Agent, Memory, Skills, and automation capabilities does not require connecting wallets, purchasing tokens, or understanding Solana. Meanwhile, Paradigm Capital, Psyche, distributed training, and the Crypto AI community still exist in the product backend. This forms a product form that can be summarized as "Crypto-native in organization, crypto-invisible in product"—retaining the most valuable parts of Crypto at the organizational level (capital, global community, user sovereignty, and coordination capabilities) while removing parts that easily hinder mainstream adoption at the product level (wallets, tokens, speculative narratives, and on-chain operational friction).

Why OpenClaw Rejects Crypto, and Why Hermes Hides Crypto

The apparent opposition between OpenClaw and Hermes on the Crypto issue is not an ideological struggle of "rejection" and "embrace." From the product outcomes, both reflect orientations of open source, user control, and reducing platform lock-in; the difference is that Nous further applies the cryptocurrency economic mechanism for distributed training coordination, while OpenClaw primarily achieves user sovereignty through a Local-first architecture:

• OpenClaw (Local-first Sovereignty): Resists financial speculation and defends "local priority" sovereignty. Early encounters with fake coin scams prompted a "zero tolerance" towards Crypto. It defends user sovereignty in a non-blockchain way through pure open source and local operations, resolutely rejecting financialization at the product level.
• Hermes/Nous (Cryptoeconomic Sovereignty): Engineering-oriented, with Crypto serving only as an underlying coordination tool. Introducing blockchain is a pragmatic choice to address engineering challenges (such as the Psyche network utilizing Solana to coordinate heterogeneous computing power), rather than constructing a financial narrative aimed at end-users.

Hermes's Advanced Model—From Personal Agent to Task Housekeeper

This section hopes to answer a more fundamental question: When Claude Code and Codex can already complete most professional execution tasks with high quality, what is the reason for the existence of Hermes as an independent product?

• Mode A: Direct cooperation type (limited gains): Users are used to manually generating prompts in LLM and handing over execution, manually transporting results and reviewing them. Although the quality of the output is high, all project management and multi-Agent coordination tasks must still be assumed. For such hands-on users, Hermes's automation is seen as "an intermediate layer that adds opacity," failing to effectively lighten their load.
• Mode B: Delegated management type (obvious gains): Users treat Hermes as the permanent control layer, only issuing final goals. Hermes is responsible for task decomposition, delegating sub-tasks, tracking GitHub/CI status, and automatically triggering rework. Community practices (like oh-my-hermes) show that Hermes's core value is precisely to replace cumbersome cross-Agent coordination and project management tasks.

Under this framework, Hermes and Claude Code/Codex are not in a substitution relationship, but a layered one: the latter provides third-layer execution quality, while the former offers second-layer continuity, cross-session state, and cross-Agent coordination. The value of Hermes is not evenly distributed among all users, but may be highly concentrated among advanced user groups engaged in cross-Agent, cross-system, and long-term asynchronous tasks. This judgment is more precise than a general statement that "the second mentality of personal Agents has formed," and is more suitable for guiding commercialization and product priorities.

▲ Figure 2 · Full view of Hermes Agent's technical architecture (user entrance → Gateway → control core → Provider layer → execution layer → orchestration layer → status layer → governance layer)

Based on official documentation and community research, the panoramic framework of Hermes Agent's technical architecture covers the entire link from user interaction to learning governance:

• System-level support for autonomous recovery: The "control core" clearly includes Context compression, Provider fallback, and interruption state saving, providing a technical basis for fault recovery and system self-healing capability during task failures.
• Execution logic of "delegation rather than substitution": The "tool and professional execution layers" align external CLIs like Claude Code, Codex with Hermes native tools (Terminal, Browser, etc.) at the same level, confirming its positioning as a scheduling hub.
• Governance attributes of "self-evolution": The "learning, maintenance, and governance layer" includes nodes like Curator and Skill/Command Approval, indicating that its experiential accumulation has governance processes with human intervention mechanisms rather than being an entirely automated black box.

Business Model—Who Pays for "Hermes"?

Comparing its token cost directly with subscribing to Claude Code/Codex will lead to misleading conclusions. This algorithm ignores the core value of Hermes: replacing users’ hands-on project management, context transportation, and cross-Agent coordination tasks.

User Value Formula Hermes user value = time saved in manual coordination + asynchronous and unattended value + cross-system automation benefits − token and tool costs − manual intervention costs − failure and security risks

Thus, the economics of Hermes is not absolute, but highly depends on the user's "delegation depth":

• High delegation depth (economic feasibility): If Hermes can turn a task that originally required hours of manual surveillance into true unattended execution, even if the token cost is slightly higher, its overall time cost and efficiency gains are still positive.
• Low delegation depth (economic collapse): If users still need frequent intervention for error correction and firefighting, Hermes will merely become a pure token consumer and a fault amplifier.

This mechanism precisely explains why different user groups have vastly different evaluations of Hermes's economics, and also reminds us that the key to validating its business logic lies in quantifying the "unattended completion rate" and "number of manual interventions per task," rather than just comparing the unit price of model APIs.

Commercialization Foundation: Nous Portal and Hermes Cloud

Hermes Agent is open-sourced under the MIT license and is positioned as an ecosystem growth engine. The real commercialization closed loop focuses on Nous Portal, whose core value proposition is "one subscription, integrating multiple API keys," covering three major modules:

• Model routing: Aggregating 252 models (providing inference through OpenRouter and direct Provider connections).
• Tool Gateway: Built-in high-frequency tools like Firecrawl (web search), FAL (image generation), Browser Use (cloud browser), Modal (sandbox execution), and OpenAI Audio (TTS).
• Managed services: Out-of-the-box Hermes Cloud instances (charged daily for operation fees, excluding inference and tool calling fees).

The actual revenue of Nous heavily relies on users' usage paths, currently showing significant structural differentiation:

Open Source and Commercialization, Will Hermes Become the "Linux of the Agent World"?

The MIT open-source strategy of Hermes, while driving explosive growth, also creates structural constraints for commercialization. The self-hosted free model requires its paid version to possess irreplaceable extra value, but so far, a clear differentiated monetization path has not formed. A deeper risk lies in "value capture": If Hermes continues to be a selectable Runtime widely integrated by cloud vendors, it may repeat the classic dilemma of Linux or K8s, where core commercial value is captured by cloud vendors providing computing power and hosting. The MIT license, while exchanging for ecosystem prosperity, also means relinquishing absolute control over distribution channels. As long as users can freely choose "self-hosting + proprietary API" or "third-party cloud deployment," the vast amount of usage cannot be forcibly converted into direct revenue, leaving Nous facing a severe test of "ecological elevation" mismatching with "actual commercial returns."

Agent Ecological Niche—Personal Housekeeper, Professional Tools, and Big Tech Claw Tripartite Pattern

OpenClaw, Hermes, and products from major companies like Claude Code and Codex have significant differences in target users and core propositions, belonging to different niche tracks. To clarify the current market landscape, the core competitive matrix for AI Agents is as follows:

Hermes does not pursue the mass market but precisely targets four categories of high-density Power Users, forming the cornerstone of its phenomenal dissemination:

• Self-hosted and infrastructure players: Familiar with VPS/Docker/SSH, viewing Hermes as a natural control layer for existing infrastructure.
• Multi-model arbitrageurs: Rejecting single vendor lock-in, accustomed to dynamically scheduling cutting-edge or local models based on tasks.
• Multi-Agent coordinators: Urgently needing to automate orchestration of complex, cross-platform, and cross-tool workflows.
• Open-source and Crypto AI community: Deeply resonating with the ideals of user sovereignty and decentralization, in sync with Nous’s organizational culture.

This type of user base, while small, possesses very high token consumption, code contribution, and technical evangelism capabilities, serving as the core engine driving early word-of-mouth dissemination.

Claude Code/Codex: Both Supplier and Threat

#Short-term Symbiosis: Raising Execution Limits

In practical workflows, Hermes acts as the control layer, calling Codex (code implementation) and Claude Code (architecture and review) through a delegation mechanism. The stronger the underlying professional Agents, the higher the task complexity limits Hermes can deliver, forming a symbiotic relationship where "Hermes is responsible for routing and acceptance, while professional Agents are responsible for execution."

#Long-term Risk of Absorbing Hermes's Independent Value

Model vendors are accelerating penetration into the control layer, posing a threat closer than anticipated. Anthropic's Claude Managed Agents already support multi-Agent parallel orchestration; OpenAI has clearly positioned Codex App as a “command center for agents,” supporting multi-Agent parallelization, automation, and long-term backend operations. This indicates that Codex's multi-Agent control capabilities within software engineering boundaries are relatively mature, even surpassing Hermes in some aspects, and is no longer just a "bottom layer executor."

Currently, Hermes possesses advantages in personal control planes across channels, models, and projects; however, Codex already has strong task ownership and multi-Agent management capabilities within the software engineering domain, and this competitive edge may be stronger than Hermes’s in that boundary. The core competitive question is whether Hermes can prioritize the sedimentation of users' project statuses, approval rules, Skills, Memory, and cross-Agent workflows above model vendors, forming assets users are reluctant to migrate from, or if it will ultimately get absorbed by model-native products as standard features?

Big Tech Company Agent Route Choices

To discuss how major companies respond to the wave of personal Agents, one must first clarify their product boundaries: on one hand, there are resident Agent hosting aimed at individuals (such as Tencent QClaw, ByteDance ArkClaw), and on the other, there are general work Agents aimed at office/enterprise (like WorkBuddy, Trae) with vastly different positioning:

• Big Tech Claw Route: Reduces barriers through one-click deployment, preset templates, and local ecosystem integrations. However, the deeper chasm lies in distrust of platform incentives: regardless of how many external models are supported, users inherently believe the ultimate goal is to channel them into their own cloud and model system.
• Hermes Runtime Integration: ByteDance's ArkClaw and Tencent Cloud have formally integrated Hermes Agent as an optional plugin or exclusive template into their cloud control panels, establishing a clear multi-Runtime strategy: large companies retain their own cloud hosting, billing, security, and enterprise-level management foundation while viewing Hermes as a pluggable advanced component, achieving complementary coexistence between open-source ecology and commercial cloud platforms.
• Transition of General Office Agents: Currently, major companies are shifting core resources from Claw to general office Agent platforms with clear demands, easy acceptance, and direct monetization (like WorkBuddy). Such tasks can be deeply integrated with their ecosystems like WeChat, DingTalk, and Feishu and converted into revenue.

Hermes’s Insights for Crypto AI

Web3 did not directly make Hermes a smarter Agent but provided Nous with a capital structure, organization methods, early high-intensity user pools, and sources of values different from traditional AI startups. Hermes has at least proposed a more mature path for Crypto AI: making Crypto an organizational and infrastructural aspect rather than a product interface that users must face.

Hermes has completed the transition from a Crypto AI research brand to a global open-source Agent product, establishing large-scale attributable inference activities and a clear second mentality—but this mentality is still concentrated in the OpenRouter ecosystem and among global developer circles, not transforming into a comprehensive overtaking of OpenClaw in terms of GitHub stars or overall community size. It lacks exclusive technologies that OpenClaw cannot replicate; instead, it completed a worthy iteration of a challenger product by precisely inheriting high-intensity users and establishing "delegable" and "self-evolving."

• Insight One: Crypto can serve as an "organizational operating system" rather than a product function: The true value of Web3 can be reflected in capital structure, early high-intensity user pools, and foundational values, without necessarily exposing it as wallet or token interaction. Achieving "Crypto-native at the organizational level, crypto-invisible at the product level" is an effective strategy balancing innovation drive and user experience.
• Insight Two: Decentralized infrastructure must anchor demand-side entries to form closed loops: Purely supply-side distributed training networks (such as DisTrO, Psyche) struggle to prove their commercial value without real user entry and execution data support. Hermes Agent is the key validation for Nous's leap from basic computing infrastructure to real demand sides.
• Insight Three: Moats can be built on "Delegation Trust" rather than solely on "model capabilities": The differentiation of personal Agents may not stem from stronger single execution capabilities but rather from whether "users dare to entrust long-term responsibilities to it." This soft trust asset is often overlooked yet has significant barriers in Crypto AI projects.
• Insight Four: Relationships with cloud giants are not zero-sum games but ecological complements: Large companies integrating Hermes as an optional Runtime prove that open-source Runtimes and corporate control planes can coexist. For entrepreneurs, "being integrated" is a viable commercialization path; however, caution is needed regarding the risk of core values being captured by cloud vendors' hosting layers.
• Insight Five: The competitive endgame will shift from "single execution capability" to "task ownership and trust accumulation": In the future, the most valuable assets may not necessarily be the strongest models at the execution layer but rather a "supervisory control system" that can accept final goals, maintain long-term context, intelligently schedule professional executors, and allow users to confidently delegate responsibilities.

OpenClaw established "personal ownership of Agents" as a clear product category; Hermes has advanced "long-term delegation of Agents" into a more systematic product direction through persistent state, task recovery, evidence acceptance, multi-model supply, and professional Agent delegation. The real test will be whether users are still willing to entrust the final goals and long-term trust to this open Runtime from a Web3 background—and continue to pay for it—when Claude Code and Codex's control capabilities in the software engineering boundaries continue to strengthen, and when large cloud platforms make multi-Runtime integrations smoother.