A command is a shortcut: you type a slash and a word, and the AI knows what kind of work you expect. But Claude and ChatGPT do not use them for the same thing, and mixing them up is the fast lane to a mediocre result.
Claude's cover a bit of everything: reviewing code, tidying a project, summarising, documenting, managing the session. The ChatGPT ones collected here do exactly one thing, create images: they set the kind of visual representation you want before you describe the scene.
A bit of everything
Slash commands for Claude Chat, Claude Cowork and Claude Code. They cover development, documentation, data, content and productivity: from clearing a session's context to auditing an entire repository.
Images only
Visual codes such as /xray, /blueprint or /mockup. They do not do the work for you: they tell ChatGPT which visual register you want before it starts drawing.
Not exactly. They are conventions that caught on among people who generate images, not a fixed product menu. They work because they steer the model, not because there is a button behind them. To see the shortcuts actually enabled on your account, type a slash in a new chat.
Claude Code's slash commands are real: they execute something. Outside Claude they mean nothing. Visual codes, by contrast, are guiding text: you can try them in any image generator, though the result will vary.
None of them. A code like /xray does not produce a real X-ray: it generates a visual interpretation. The more specific the rest of the prompt (what must appear, in what style, in what format and what stays out), the less room for interpretation you leave.
Yes. Claude's commands change with every release and the dictionary is reviewed alongside them. Visual codes are added when one brings something genuinely different, not to pad the list.
Custody, risk control and AI strategy bots in a single dashboard. Backtests, calibrated parameters and real-time execution, with the whole process in plain sight and available on mobile too.
