ai·wholetech·com ▶Now Playing — the videos worth your time, hand-pickedWatch →
⚡ Built Agents Firstwhy →|AGENTS.mdllms.txtopenapiMCP
⌁ the AI tooling hub on the WholeTech network

Five AI tools, one hub, in plain English.

Deep, evergreen guides to the five AI tools the WholeTech network actually uses — Claude, OpenAI Codex, Google Gemini, xAI's Grok, and Hermes. No hype, no jargon: how each one works, what it's best at, and how they fit together.

Each section is a full mirror of its standalone subdomain, kept in one place so the comparisons are obvious and the navigation is one click.

Updated continuously Plain language Part of the WholeTech network

Assistants

/assistants/ →

Four AI products the WholeTech network actually uses. Each is partly a model (Layer 1) and partly an agent or chat surface (Layer 3); the brand-name pages cover both.

C

Claude

/assistants/claude/

A complete, plain-language guide to Claude and the wider Anthropic toolkit — products, design workflow, integrations, and machine builds. The most-used model on the WholeTech network.

Maker: AnthropicBest at: careful reasoning, long docs, coding
→
X

Codex

/assistants/codex/

An end-to-end guide to OpenAI Codex — the CLI, the IDE extension, the cloud, GitHub @codex, AGENTS.md, approval modes, prompting, and the workflows that actually pay off.

Maker: OpenAIBest at: raw coding throughput
→
G

Gemini

/assistants/gemini/

The practical guide to Google's AI stack — the Gemini app, AI Studio, Gemini API, Workspace, Android, Search, model choices, workflows, and how Gemini fits beside Claude, Codex, and Hermes.

Maker: GoogleBest at: long context, Google integration
→
K

Grok

/assistants/grok/

The end-to-end guide to xAI's Grok and the platform formerly known as Twitter — the models (Grok 4, 4 Heavy, mini, Aurora, Imagine), every surface, every plan, the timeline algorithm, and how to actually win on X.

Maker: xAIBest at: real-time, X-native context
→

Orchestration

Not an assistant — a multi-model harness. Points at any of the four above and ties them together with files, deployments, and repeatable workflows.

H

Hermes

/hermes/

The orchestration harness around Claude, Codex, Gemini, local tools, website work, deployments, sessions, and repeatable agent workflows. The connective tissue of the network.

Maker: Nous ResearchBest at: multi-tool, multi-model orchestration
→

Device setup guides

Once you've picked a model above, the next question is "how do I run this on the machine in front of me." The OS guides answer that — six clean steps per device, plus a deep analysis page for each.

OS

OS · WholeTech

/os/

Plain-language setup guides for Windows, Mac, Linux, Synology, Android, iOS, Proxmox, and Cloud — plus a critical analysis page for each (including the optional editor route via VS Code), a four-layer tool map, and a one-page cheat sheet of every install command.

8 devices8 analyses1 tool map1 cheat sheet
→
What changed: these guides used to live at claude.wholetech.com, codex.wholetech.com, gemini.wholetech.com, grok.wholetech.com, hermes.wholetech.com, and os.wholetech.com. Those URLs now 301-redirect here, so old bookmarks land in the right place automatically. The four chat/coding assistants are grouped under /assistants/; Hermes (the harness) and the OS guides each get their own top-level path.

Frequently asked questions

Straight answers to the questions people ask before picking an AI tool.

Which AI tool should I use — Claude, Codex, Gemini, or Grok?
It depends on the job. Reach for Claude for careful reasoning, long documents, and coding; Codex for raw coding throughput in the terminal and IDE; Gemini for very long context and tight Google integration; and Grok for real-time, X-native context. Most people end up using two or three depending on the task rather than crowning a single winner.
What is each of these AI tools best at?
Claude leans toward careful reasoning, long docs, and coding; Codex is built for high-volume coding through its CLI and IDE extension; Gemini shines on long context and Google integration; and Grok is strongest on real-time, X-native information. Hermes is not an assistant at all but an orchestration harness that ties the others together. The assistants overview breaks each one down in plain language.
Is Claude better than ChatGPT or Gemini for coding?
Claude is one of the strongest coding models and the most-used one across the WholeTech network, but "better" depends on the task — OpenAI's Codex is purpose-built for raw coding throughput, and Gemini's long context helps on large codebases. Rather than crown one winner, the hub covers all of them so you can match the tool to the job. Compare the Claude guide and the Codex guide directly.
What's the difference between Claude and Codex?
Claude, made by Anthropic, is a general-purpose assistant strong at reasoning, long documents, and coding across both chat and agent surfaces. Codex, from OpenAI, is a coding-focused product — a CLI, IDE extension, and cloud agent built around raw coding throughput, GitHub, and AGENTS.md files. They overlap on code, but Claude covers a wider range of work; full details are in the Claude and Codex guides.
What is Hermes, and how is it different from the four assistants?
Hermes is not a chat assistant — it is an orchestration harness that points at Claude, Codex, or Gemini and ties them together with files, deployments, sessions, and repeatable workflows. Think of the four assistants as the engines and Hermes as the connective tissue that runs them across a whole project. Made by Nous Research, it is best at multi-tool, multi-model orchestration.
How do I get started with these AI tools?
Pick a model from the guides above, then use the OS setup guides for six clean steps to run it on Windows, Mac, Linux, Synology, Android, iOS, Proxmox, or Cloud. If you are newer to working with agents, the plain-language agent primer at fableguide.com is a good on-ramp. From there, each tool's guide walks through the workflows that actually pay off.
How do these five tools fit together across a network of sites?
No single tool does everything, so the WholeTech network uses all five and lets each play to its strengths. The network page shows how the guides, tools, and sites connect, and how orchestration with Hermes keeps the work repeatable across many machines and sites. The goal is picking the right tool per task, not standardizing on one.
Live
◐ Theme
Related reading: Build an autonomous agent with Claude Fable 5 · FableGuide — the Claude Fable 5 field guide · Claude on the WholeTech network · Local-first AI: keep your data on your own machine · A wearable memory aid that keeps your day private, at home