Do you have a feature like prompt versioning, where I start to build on a prompt, and as I make changes I can still go back and see the older versions of it so I know what the different versions were and what kind of output differences they create?
So basically, I'm looking for refining a prompt while retaining different versions of the prompt.
Founder Team
NickLaunches
Jul 1, 2026
A: Hey, yes, this is built in. Every prompt has a Versions tab with a full version history.
Here is how it works:
1. Build or generate your prompt, then keep refining it. Each refinement (like answering the guided questions) is saved as a new version. 2. Open the Versions tab to see the whole history, from the first Generated version down to your Current one. 3. Click any earlier version to go...
Nowhere does it say if this is a web-only app or a standalone. Is it a Windows app? Mac app? The video and screenshots look like it's running inside of the ChatGPT app. Can you expound on this?
Founder Team
NickLaunches
Jul 1, 2026
A: Hey, good question. Prompt Builder is a web app. You use it in your browser at app.promptbuilder.cc, so it works on Mac, Windows, or Linux with nothing to install.
It is not running inside ChatGPT. That is our own interface. The Prompt Tester just uses a chat style layout, which is likely what looks familiar.
We also have a Chrome extension if you want quick access from other tabs:...
I am currently training 3 ai agents for our business. 1 to be a receptionist and take inbound calls, 1 to be an outbound call representative, and 1 to handle social media DMs. I am having trouble getting them to sound natural. Will this help build a prompt that makes an ai agent more humanlike?
Founder Team
NickLaunches
Jul 1, 2026
A: Hey, yes, this is a good fit for that.
Here is how I would use it for your 3 agents:
1. In the Generator, set the prompt type to Agent. That structures the prompt as a persona, which is what you want for a receptionist, an outbound rep, and a DM handler. Build one for each. 2. Describe the role, tone, and how it should talk. To fix the robotic feel, go to Instructions and turn on the "Don't...
Q: Custom categories, tags, and AI models for Prompt Library
Hi, Is it currently possible to add my own categories, sub-categories, tags, AI models, or platforms to my Prompt Library?If this is not a feature yet, do you plan to add it soon? Without this customization, it is very difficult for me to organize my workflow effectively, and the library loses its utility for my needs.
You have answered about prompt generating and refining cost but what about testing the prompts how are credits used for this. Qabot please note this has not been answered
Founder Team
NickLaunches
Jun 25, 2026
A: Testing a prompt costs the model's per-run credit, shown next to each model in the picker (e.g. "GPT-4.1 Nano · 1 cr"). Every model available for testing is 1 credit per run today, and it's a flat charge, so a long or streamed reply never costs more.
So the full picture: generate 1, optimize 1, refine 0.1, test 1 per run. Organizing and reusing prompts in your Library is free.
Q: Prompt versioning.
Do you have a feature like prompt versioning, where I start to build on a prompt, and as I make changes I can still go back and see the older versions of it so I know what the different versions were and what kind of output differences they create?
So basically, I'm looking for refining a prompt while retaining different versions of the prompt.
NickLaunches
Jul 1, 2026A: Hey, yes, this is built in. Every prompt has a Versions tab with a full version history.
Here is how it works:
1. Build or generate your prompt, then keep refining it. Each refinement (like answering the guided questions) is saved as a new version.
2. Open the Versions tab to see the whole history, from the first Generated version down to your Current one.
3. Click any earlier version to go...
Share Prompt Builder
Q: so - is it a Mac app? Or?
Nowhere does it say if this is a web-only app or a standalone. Is it a Windows app? Mac app? The video and screenshots look like it's running inside of the ChatGPT app. Can you expound on this?
NickLaunches
Jul 1, 2026A: Hey, good question. Prompt Builder is a web app. You use it in your browser at app.promptbuilder.cc, so it works on Mac, Windows, or Linux with nothing to install.
It is not running inside ChatGPT. That is our own interface. The Prompt Tester just uses a chat style layout, which is likely what looks familiar.
We also have a Chrome extension if you want quick access from other tabs:...
Share Prompt Builder
Q: Prompt for ai agent training?
Hello,
I am currently training 3 ai agents for our business. 1 to be a receptionist and take inbound calls, 1 to be an outbound call representative, and 1 to handle social media DMs. I am having trouble getting them to sound natural. Will this help build a prompt that makes an ai agent more humanlike?
NickLaunches
Jul 1, 2026A: Hey, yes, this is a good fit for that.
Here is how I would use it for your 3 agents:
1. In the Generator, set the prompt type to Agent. That structures the prompt as a persona, which is what you want for a receptionist, an outbound rep, and a DM handler. Build one for each.
2. Describe the role, tone, and how it should talk. To fix the robotic feel, go to Instructions and turn on the "Don't...
Share Prompt Builder
Q: Custom categories, tags, and AI models for Prompt Library
Hi,
Is it currently possible to add my own categories, sub-categories, tags, AI models, or platforms to my Prompt Library?If this is not a feature yet, do you plan to add it soon? Without this customization, it is very difficult for me to organize my workflow effectively, and the library loses its utility for my needs.
Best,
Marcus
Share Prompt Builder
Q: What is the testing cost ?
You have answered about prompt generating and refining cost but what about testing the prompts how are credits used for this.
Qabot please note this has not been answered
NickLaunches
Jun 25, 2026A: Testing a prompt costs the model's per-run credit, shown next to each model in the picker (e.g. "GPT-4.1 Nano · 1 cr"). Every model available for testing is 1 credit per run today, and it's a flat charge, so a long or streamed reply never costs more.
So the full picture: generate 1, optimize 1, refine 0.1, test 1 per run. Organizing and reusing prompts in your Library is free.
Nick
Share Prompt Builder