{"id":30830,"date":"2026-05-28T16:32:12","date_gmt":"2026-05-28T14:32:12","guid":{"rendered":"https:\/\/contabo.com\/blog\/?p=30830"},"modified":"2026-05-29T12:22:30","modified_gmt":"2026-05-29T10:22:30","slug":"hermes-agent-vs-openclaw-paperclip-and-the-best-open-source-ai-agents-in-2026","status":"publish","type":"post","link":"https:\/\/contabo.com\/blog\/hermes-agent-vs-openclaw-paperclip-and-the-best-open-source-ai-agents-in-2026\/","title":{"rendered":"Hermes Agent vs OpenClaw, Paperclip, and the Best Open-Source AI Agents in 2026"},"content":{"rendered":"\n
\"Hermes<\/figure>\n\n\n\n

Article Summary:<\/strong> Hermes Agent is a model-agnostic open-source agent runtime from Nous Research. OpenClaw is a Node.js personal-assistant gateway with channels for WhatsApp, Telegram, Slack, and Discord. Paperclip is a multi-agent orchestrator that can call Hermes Agent as a managed worker. Hermes 4 ships in 70B and 405B (Aug 2025); Hermes 3 still wins on cheap VPS deployments with its 3B and 8B sizes.<\/p>\n\n\n\n

\n\n\n\n

The open-source AI agent space looks different in 2026 than it did a year ago. Hermes vs OpenClaw is no longer a niche debate, and the Hermes 3 versus Hermes 4 model choice now matters as much as the agent runtime sitting in front of it. This article compares Hermes Agent against OpenClaw and Paperclip, then walks through Hermes 3 versus Hermes 4 so you can pair the right model with the right runtime on a Contabo VPS.<\/p>\n\n\n\n

Open-Source AI Agents in 2026: A Quick Landscape<\/h2>\n\n\n\n

Three projects dominate the self-hosted conversation. Hermes Agent<\/strong> (the Nous Research open-source agent runtime) targets developers who want a small, scriptable process running on a single VPS, with bring-your-own-model support across Nous Portal, OpenRouter, OpenAI, NovitaAI, NVIDIA NIM, Hugging Face, and local endpoints. OpenClaw<\/strong> aims at the personal-assistant axis: a Node.js gateway that fans out to messaging channels like WhatsApp, Telegram, Slack, Discord, iMessage, Signal, Teams, and Matrix, with skills and tool-use built in. Paperclip AI<\/strong> Hermes integrations sit a layer up, since Paperclip is a multi-agent orchestrator that delegates to worker agents, and Hermes Agent ships an official hermes_local<\/code> adapter for use as a managed employee inside a Paperclip company. Underneath all three sit the actual language models, and the pairing you pick decides cost, VRAM, and how much glue code you write yourself.<\/p>\n\n\n\n

Hermes Agent vs OpenClaw: Side-by-Side<\/h2>\n\n\n\n

The shortest answer on Hermes Agent vs OpenClaw is this: Hermes Agent optimizes for server-side footprint and scripting, OpenClaw optimizes for multi-channel personal-assistant reach. The OpenClaw vs Hermes AI agent comparison below covers the six dimensions that change deployment decisions.<\/p>\n\n\n\n

Dimension<\/th>Hermes Agent<\/th>OpenClaw<\/th><\/tr><\/thead>
Primary form factor<\/td>Headless daemon, CLI, HTTP API on port 8642<\/td>Node.js gateway (npm\/pnpm\/bun), runs on macOS\/Linux\/Windows via WSL2<\/td><\/tr>
Default model<\/td>Model-agnostic; supports Nous Portal, OpenRouter, OpenAI, NovitaAI, NVIDIA NIM, HF, local<\/td>Bring-your-own, Anthropic\/OpenAI\/local models<\/td><\/tr>
Plugin \/ tool ecosystem<\/td>Smaller, code-first tool definitions<\/td>Skills + channel adapters (WhatsApp, Telegram, Slack, Discord, iMessage, Signal, Teams, Matrix)<\/td><\/tr>
Minimum VPS profile<\/td>4 GB RAM, 2 vCPU for small models<\/td>8 GB RAM, 4 vCPU recommended<\/td><\/tr>
GPU requirement<\/td>Optional for 7B-13B, required for 70B+<\/td>Optional, depends on model backend<\/td><\/tr>
License<\/td>Open source<\/td>Open source<\/td><\/tr><\/tbody><\/table><\/figure>\n\n\n\n

When to Choose Hermes Agent over OpenClaw<\/h3>\n\n\n\n

Pick the Hermes Agent vs OpenClaw side when the deployment is server-first and you care about predictable cost:<\/p>\n\n\n\n

    \n
  • You run the agent on a VPS without a desktop, and you reach it over SSH or HTTP.<\/li>\n\n\n\n
  • You want a small process that pairs cleanly with Hermes 3 or Hermes 4 without extra adapters.<\/li>\n\n\n\n
  • Your toolchain is code, not GUI: n8n flows, cron jobs, shell scripts, or your own backend.<\/li>\n\n\n\n
  • You need to fit inside 4-8 GB of RAM on a budget VPS instance.<\/li>\n\n\n\n
  • You want the simplest possible upgrade path between Hermes model versions.<\/li>\n<\/ul>\n\n\n\n

    When to Choose OpenClaw over Hermes Agent<\/h3>\n\n\n\n

    Pick OpenClaw when the user-facing surface matters more than the server footprint:<\/p>\n\n\n\n

      \n
    • You want a personal assistant reachable from WhatsApp, Telegram, Slack, Discord, iMessage, Signal, Teams, or Matrix.<\/li>\n\n\n\n
    • You want a Node.js codebase you can extend with custom skills.<\/li>\n\n\n\n
    • The deployment is a single workstation or a small personal server, not a production fleet.<\/li>\n\n\n\n
    • You plan to mix Anthropic, OpenAI, and local models behind one assistant.<\/li>\n<\/ul>\n\n\n\n

      Migrating from OpenClaw to Hermes Agent<\/h3>\n\n\n\n

      OpenClaw vs Hermes migration is mostly about exporting prompts and tool definitions, then re-pointing them at the Hermes Agent HTTP endpoint. The migration command below assumes you already exported an OpenClaw workspace bundle.<\/p>\n\n\n\n

      hermes-agent import --from openclaw --bundle .\/openclaw-export.zip --target http:\/\/127.0.0.1:8642<\/code><\/pre>\n\n\n\n
      After import, re-run your test prompts against the new endpoint before retiring the OpenClaw install.<\/p>\n\n\n\n
      Hermes Agent vs Paperclip: Single Agent or Multi-Agent Org?<\/h2>\n\n\n\n
      The Paperclip AI or Hermes question is not a head-to-head: Paperclip is an orchestrator, Hermes Agent is a worker, and the two are designed to compose via the official hermes-paperclip-adapter<\/code>. Use Hermes Agent alone when one well-prompted process can handle the job: a single inbox triage flow, one code-review bot, one customer-support assistant. Use Paperclip when you need several specialized agents that pass work between each other, each with its own prompt and tools. In a Paperclip topology, Hermes Agent is registered as a hermes_local<\/code> managed employee inside a Paperclip company, usually the cheap, fast worker behind one or more roles, with a larger model reserved for the planner. The right question is rarely Hermes Agent or Paperclip; it is whether the workload is one agent’s job or a team’s.<\/p>\n\n\n\n
      Other Hermes Agent Alternatives Worth Knowing<\/h2>\n\n\n\n
      If neither Hermes Agent nor OpenClaw nor Paperclip fits, three other names come up often when people ask what the best Hermes AI alternative is. The Hermes AI agent framework conversation includes lighter scripting libraries and heavier orchestrators alike.<\/p>\n\n\n\n
      Alternative<\/th>Best for<\/th>Trade-off<\/th><\/tr><\/thead>
      LangGraph<\/td>Graph-shaped, stateful agent workflows in Python<\/td>Steeper learning curve than a single-process agent<\/td><\/tr>
      AutoGen<\/td>Multi-agent conversation patterns with role play<\/td>Conversation-driven model still maturing in production tooling<\/td><\/tr>
      CrewAI<\/td>Small teams of role-based agents with clear roles and tasks<\/td>Less flexible for complex non-linear workflows than LangGraph<\/td><\/tr><\/tbody><\/table><\/figure>\n\n\n\n
      Treat the table as a starting list, not a ranking. The right pick depends on whether you want a library, a runtime, or an orchestrator.<\/p>\n\n\n\n
      Hermes 3 vs Hermes 4: Full Model Comparison<\/h2>\n\n\n\n
      Hermes 3 AI and Hermes 4 AI are the two Nous Research model generations to choose between in 2026. The Hermes AI model decision sits underneath every agent choice above: a great runtime cannot rescue an underpowered model, and an oversized one wastes VPS budget. The Hermes 3 AI model line is older, broader, and ships in 3B, 8B, 70B, and 405B sizes. The Hermes 4 AI model is newer (released August 2025), ships in 70B and 405B, and adds hybrid-mode reasoning.<\/p>\n\n\n\n
      Hermes 3 AI Model: Sizes, Variants, and Best Use Cases<\/h3>\n\n\n\n
      The Hermes 3 AI model ships in 3B, 8B, 70B, and 405B sizes, all fine-tuned on Llama 3.1 or 3.2. The Hermes 13B AI label refers to the legacy Nous-Hermes-Llama2-13B fine-tune, not a Hermes 3 size. The Nous Hermes 3 AI model brand also includes uncensored variants for research workloads. The table below gives a rough sizing reference. Confirm exact figures against the official Nous Research model card before sizing your VPS.<\/p>\n\n\n\n
      Model<\/th>VRAM (GPU)<\/th>CPU-only feasible?<\/th>Best for<\/th><\/tr><\/thead>
      Hermes 3 3B<\/td>~3-6 GB<\/td>Yes<\/td>Edge and budget VPS chat<\/td><\/tr>
      Hermes 3 8B<\/td>~16 GB FP16, ~8 GB 4-bit<\/td>Yes, slow<\/td>Cheap VPS chat and tool-use<\/td><\/tr>
      Hermes 3 70B<\/td>~140 GB FP16, ~40 GB 4-bit<\/td>No<\/td>Higher-quality reasoning on GPU servers<\/td><\/tr>
      Hermes 3 405B<\/td>~810 GB FP16, ~430 GB FP8<\/td>No<\/td>Research and benchmarking on multi-GPU rigs<\/td><\/tr>
      Nous-Hermes-Llama2-13B (legacy Hermes 13B AI)<\/td>~26 GB FP16, ~8 GB 4-bit<\/td>Marginal<\/td>Legacy projects still pinned to Llama 2<\/td><\/tr>
      Hermes 3 uncensored AI model variants<\/td>Same as base size<\/td>Same as base<\/td>Red-team research, fiction, policy testing<\/td><\/tr><\/tbody><\/table><\/figure>\n\n\n\n
      Hermes 4 AI Model: What Changed and Who Should Upgrade<\/h3>\n\n\n\n
      The Hermes 4 AI model line ships in 70B and 405B sizes, released August 26, 2025, and adds hybrid-mode reasoning over Hermes 3. Open-source AI model releases for Hermes 4 are tracked on Nous Research and Hugging Face. Upgrade when your bottleneck is answer quality, not infrastructure cost. Stay on Hermes 3 when you are RAM-bound, VRAM-bound, or need the 3B or 8B sizes.<\/p>\n\n\n\n
      Older Hermes Models: Nous Hermes 2, Chronos-Hermes, OpenHermes<\/h3>\n\n\n\n
      Several earlier Hermes variants still ship in tutorials and community repos:<\/p>\n\n\n\n
        \n
      • Nous Hermes 2: the predecessor line to Hermes 3, with the Mixtral 8x7B DPO release from January 2024 still in active community use.<\/li>\n\n\n\n
      • Chronos-Hermes: a 13B merge by Austism focused on long-context narrative and roleplay, often shipped via TheBloke quantizations.<\/li>\n\n\n\n
      • OpenHermes: an open release based on Mistral 7B fine-tunes, often used as a teaching example.<\/li>\n\n\n\n
      • Nous-Hermes-Llama2-13B: a Llama 2 era 13B fine-tune from July 2023 that still appears in legacy projects.<\/li>\n\n\n\n
      • DeepHermes 3: a preview reasoning-focused fine-tune of Llama 3 8B from February 2025.<\/li>\n<\/ul>\n\n\n\n
        Treat these as legacy unless a specific workflow pins to them. New deployments should start on Hermes 3 or Hermes 4.<\/p>\n\n\n\n
        Pairing Hermes Agent with Hermes 3 or Hermes 4 Locally<\/h2>\n\n\n\n
        To pair Hermes Agent with a local Hermes AI model on a VPS, work through the steps below. They assume Ubuntu 22.04, root SSH access, and that the agent will run as a systemd service:<\/p>\n\n\n\n
          \n
        1. Provision a VPS sized for the target model: 4 GB RAM for Hermes 3 3B, at least 16 GB for Hermes 3 8B and above.<\/li>\n\n\n\n
        2. Install the Hermes Agent runtime and confirm the service listens on API_SERVER_PORT<\/code> 8642. Hermes Agent is also available as a free 1-click Add-On for your server, which saves you some of the setup work.<\/li>\n\n\n\n
        3. Pull the chosen Hermes 3 or Hermes 4 weights from the official Nous Research source.<\/li>\n\n\n\n
        4. Register the model in the agent config and run a smoke-test prompt against the local endpoint.<\/li>\n\n\n\n
        5. Wire the endpoint into your tools (n8n, scripts, or Paperclip) using the same HTTP API.<\/li>\n<\/ol>\n\n\n\n
          Decision Matrix: Pick Your Stack in 60 Seconds<\/h2>\n\n\n\n
          Use the matrix below as a starting point. Pick the row that matches your primary use case, then read across for the recommended agent and model pairing.<\/p>\n\n\n\n
          Use case<\/th>Agent<\/th>Model<\/th><\/tr><\/thead>
          Single-VPS chatbot or tool-use bot<\/td>Hermes Agent<\/td>Hermes 3 3B or 8B<\/td><\/tr>
          Personal assistant across messaging channels<\/td>OpenClaw<\/td>Hosted Claude or OpenAI, or local Hermes 3 8B<\/td><\/tr>
          Multi-agent content pipeline<\/td>Paperclip orchestrating Hermes Agent workers<\/td>Hermes 4 70B planner, Hermes 3 8B workers<\/td><\/tr>
          High-quality reasoning on GPU server<\/td>Hermes Agent<\/td>Hermes 4 70B or Hermes 3 70B<\/td><\/tr>
          Research on uncensored variants<\/td>Hermes Agent<\/td>Hermes 3 uncensored variant<\/td><\/tr><\/tbody><\/table><\/figure>\n\n\n\n
          Why Self-Host Your Hermes Comparison Stack on Contabo<\/h2>\n\n\n\n
          Self-hosting Hermes Agent, OpenClaw, or Paperclip on a Contabo VPS or GPU server gives you a predictable monthly cost, full control over which model weights you pull, and the option to keep prompts and customer data on infrastructure you operate. The same VPS can host the agent runtime today and the next model generation a year from now, without changing vendors or paying per-token fees.<\/p>\n\n\n\n
          FAQ: Hermes Agent and Hermes Models Compared<\/h2>\n\n\n\n
          What is the difference between Hermes Agent and OpenClaw?<\/strong> 
          Hermes Agent is a small open-source AI agent runtime from Nous Research, designed to run as a headless service on a VPS and pair with any supported model provider, including the Hermes family. OpenClaw is a Node.js personal-assistant gateway that connects channels like WhatsApp, Telegram, Slack, and Discord to a chosen model backend. The OpenClaw vs Hermes choice usually comes down to whether you need a server-side agent or a multi-channel personal assistant.<\/p> <\/div> 
          Is Hermes 4 better than Hermes 3?<\/strong> 
          Hermes 4 AI is the newer Nous Research generation, released August 2025 in 70B and 405B sizes with hybrid-mode reasoning. Whether it is better for you depends on workload. Hermes 4 wins on answer quality at the top end, while the Hermes AI model decision often lands on Hermes 3 when VPS cost, smaller VRAM, the 3B or 8B sizes, or uncensored variants matter more.<\/p> <\/div> 
          What is the best Hermes AI agent alternative in 2026?<\/strong> 
          There is no single best Hermes AI alternative. The right Hermes AI agent alternative depends on shape. For multi-agent orchestration, Paperclip is the closest peer. For graph-shaped Python workflows, LangGraph fits well. For multi-channel personal assistants, OpenClaw is the closest match. Pick by deployment shape first, then by model fit.<\/p> <\/div> 
          Can I run Hermes 3 13B or 70B on a VPS?<\/strong> 
          Yes for the legacy Hermes 13B AI fine-tune (Nous-Hermes-Llama2-13B) on a sufficiently sized VPS, with the practical floor around 16 GB RAM for CPU-only 4-bit inference or a small GPU plan for usable speeds. The Hermes 3 70B AI model is a different class: plan for a GPU server with roughly 140 GB of VRAM at FP16 or ~40 GB at 4-bit quantization, not a standard VPS. For the Hermes 3 AI model in general, match the size to your RAM and VRAM budget before picking the variant.<\/p> <\/div> 
          Does Hermes Agent work with Paperclip?<\/strong> 
          Yes. Paperclip AI Hermes integrations are first-class through the official hermes-paperclip-adapter<\/code>, which registers Hermes Agent as a hermes_local<\/code> managed employee inside a Paperclip company. The Hermes AI agent role inside Paperclip is usually a fast, cheap worker behind a more capable planner, though you can also run Hermes Agent as the planner if you prefer.<\/p> <\/div> 
          Where is the Hermes Agent GitHub repository?<\/strong> 
          The Hermes Agent GitHub repository is published by the Nous Research organization at https:\/\/github.com\/nousresearch\/hermes-agent. Confirm the exact Hermes Agent GitHub AI URL on the Nous Research website before cloning, since several community forks share similar names. Pin to a released tag rather than tracking the default branch in production.<\/p> <\/div> 
          What is an uncensored Hermes 3 AI model?<\/strong> 
          A Hermes 3 AI model uncensored variant is a release of Hermes 3 with reduced refusal behavior, intended for research, red-teaming, and creative use cases. The Hermes 3 uncensored AI model variants share the base architecture of the standard Hermes 3 line, so VRAM and CPU profiles match the equivalent base size. Use them with the same care you would apply to any uncensored model, and review your hosting provider’s acceptable-use policy first.<\/p> <\/div> <\/div>\n\n\n\n
          <\/p>\n","protected":false},"excerpt":{"rendered":"
          A side-by-side look at Hermes Agent, OpenClaw, and Paperclip, with a full Hermes 3 vs Hermes 4 model comparison and a quick decision matrix for picking your 2026 stack.<\/p>\n","protected":false},"author":63,"featured_media":30855,"comment_status":"closed","ping_status":"closed","sticky":false,"template":"","format":"standard","meta":{"inline_featured_image":false,"_uag_custom_page_level_css":"","site-sidebar-layout":"default","site-content-layout":"","ast-site-content-layout":"default","site-content-style":"default","site-sidebar-style":"default","ast-global-header-display":"","ast-banner-title-visibility":"","ast-main-header-display":"","ast-hfb-above-header-display":"","ast-hfb-below-header-display":"","ast-hfb-mobile-header-display":"","site-post-title":"","ast-breadcrumbs-content":"","ast-featured-img":"","footer-sml-layout":"","theme-transparent-header-meta":"","adv-header-id-meta":"","stick-header-meta":"","header-above-stick-meta":"","header-main-stick-meta":"","header-below-stick-meta":"","astra-migrate-meta-layouts":"set","ast-page-background-enabled":"default","ast-page-background-meta":{"desktop":{"background-color":"var(--ast-global-color-4)","background-image":"","background-repeat":"repeat","background-position":"center center","background-size":"auto","background-attachment":"scroll","background-type":"","background-media":"","overlay-type":"","overlay-color":"","overlay-opacity":"","overlay-gradient":""},"tablet":{"background-color":"","background-image":"","background-repeat":"repeat","background-position":"center center","background-size":"auto","background-attachment":"scroll","background-type":"","background-media":"","overlay-type":"","overlay-color":"","overlay-opacity":"","overlay-gradient":""},"mobile":{"background-color":"","background-image":"","background-repeat":"repeat","background-position":"center center","background-size":"auto","background-attachment":"scroll","background-type":"","background-media":"","overlay-type":"","overlay-color":"","overlay-opacity":"","overlay-gradient":""}},"ast-content-background-meta":{"desktop":{"background-color":"var(--ast-global-color-5)","background-image":"","background-repeat":"repeat","background-position":"center center","background-size":"auto","background-attachment":"scroll","background-type":"","background-media":"","overlay-type":"","overlay-color":"","overlay-opacity":"","overlay-gradient":""},"tablet":{"background-color":"var(--ast-global-color-5)","background-image":"","background-repeat":"repeat","background-position":"center center","background-size":"auto","background-attachment":"scroll","background-type":"","background-media":"","overlay-type":"","overlay-color":"","overlay-opacity":"","overlay-gradient":""},"mobile":{"background-color":"var(--ast-global-color-5)","background-image":"","background-repeat":"repeat","background-position":"center center","background-size":"auto","background-attachment":"scroll","background-type":"","background-media":"","overlay-type":"","overlay-color":"","overlay-opacity":"","overlay-gradient":""}},"footnotes":""},"categories":[1535],"tags":[4333,4330,4331,4284,4332,3722,4329,3319],"ppma_author":[1492],"class_list":["post-30830","post","type-post","status-publish","format-standard","has-post-thumbnail","hentry","category-comparisons","tag-ai-agent-comparison","tag-hermes-3","tag-hermes-4","tag-hermes-agent","tag-open-source-ai-agents","tag-openclaw","tag-paperclip","tag-self-hosted-ai"],"uagb_featured_image_src":{"full":["https:\/\/contabo.com\/blog\/wp-content\/uploads\/2026\/05\/blog-head_hermes-agent-vs.webp",1200,630,false],"thumbnail":["https:\/\/contabo.com\/blog\/wp-content\/uploads\/2026\/05\/blog-head_hermes-agent-vs-150x150.webp",150,150,true],"medium":["https:\/\/contabo.com\/blog\/wp-content\/uploads\/2026\/05\/blog-head_hermes-agent-vs-600x315.webp",600,315,true],"medium_large":["https:\/\/contabo.com\/blog\/wp-content\/uploads\/2026\/05\/blog-head_hermes-agent-vs-768x403.webp",768,403,true],"large":["https:\/\/contabo.com\/blog\/wp-content\/uploads\/2026\/05\/blog-head_hermes-agent-vs.webp",1200,630,false],"1536x1536":["https:\/\/contabo.com\/blog\/wp-content\/uploads\/2026\/05\/blog-head_hermes-agent-vs.webp",1200,630,false],"2048x2048":["https:\/\/contabo.com\/blog\/wp-content\/uploads\/2026\/05\/blog-head_hermes-agent-vs.webp",1200,630,false]},"uagb_author_info":{"display_name":"Christopher Carter","author_link":"https:\/\/contabo.com\/blog\/author\/christophercarter\/"},"uagb_comment_info":0,"uagb_excerpt":"A side-by-side look at Hermes Agent, OpenClaw, and Paperclip, with a full Hermes 3 vs Hermes 4 model comparison and a quick decision matrix for picking your 2026 stack.","authors":[{"term_id":1492,"user_id":63,"is_guest":0,"slug":"christophercarter","display_name":"Christopher Carter","avatar_url":"https:\/\/secure.gravatar.com\/avatar\/63db81672a5ce4c1e8ee39753d00251d561b5b3a9967febf1c4f662024cef00f?s=96&d=mm&r=g","0":null,"1":"","2":"","3":"","4":"","5":"","6":"","7":"","8":""}],"_links":{"self":[{"href":"https:\/\/contabo.com\/blog\/wp-json\/wp\/v2\/posts\/30830","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/contabo.com\/blog\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/contabo.com\/blog\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/contabo.com\/blog\/wp-json\/wp\/v2\/users\/63"}],"replies":[{"embeddable":true,"href":"https:\/\/contabo.com\/blog\/wp-json\/wp\/v2\/comments?post=30830"}],"version-history":[{"count":2,"href":"https:\/\/contabo.com\/blog\/wp-json\/wp\/v2\/posts\/30830\/revisions"}],"predecessor-version":[{"id":30867,"href":"https:\/\/contabo.com\/blog\/wp-json\/wp\/v2\/posts\/30830\/revisions\/30867"}],"wp:featuredmedia":[{"embeddable":true,"href":"https:\/\/contabo.com\/blog\/wp-json\/wp\/v2\/media\/30855"}],"wp:attachment":[{"href":"https:\/\/contabo.com\/blog\/wp-json\/wp\/v2\/media?parent=30830"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/contabo.com\/blog\/wp-json\/wp\/v2\/categories?post=30830"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/contabo.com\/blog\/wp-json\/wp\/v2\/tags?post=30830"},{"taxonomy":"author","embeddable":true,"href":"https:\/\/contabo.com\/blog\/wp-json\/wp\/v2\/ppma_author?post=30830"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}