ChatGPT-4o vs GPT-5 for creative communications

2 min read

A man compares two different chatgpt models side by side | Martin Potter

GenAI has reached the stage where it’s no longer a novelty, it’s a tool and one that’s moving fast from curiosity to craft. We’ve stopped asking whether it can write copy or brainstorm headlines. The more interesting question is how different models behave, and what that means for how we use them in creative work.

Lately I’ve been swapping between ChatGPT-4o and GPT-5, and it struck me that they feel like two very different colleagues in the same agency. One thrives on energy and speed, the other on patience and depth. Neither is “better.” The art lies in knowing when to bring each into the room.

The fast-talker: chatGPT-4o

4o is the quick-thinking junior who thrives in a brainstorm. It gets straight to the point, fires out ideas without ego, and keeps the room buzzing. Perfect when you need options, lists, or sparks to kick things off. In practice, 4o shines at:

Headline sprints
Campaign taglines
First-round creative directions

It also has more personality than earlier versions, which makes it a fun sparring partner. The catch is it can fall in love with every idea it produces.

The deep listener: chatGPT-5

GPT-5 feels like the strategist who waits until the noise settles, then reframes the problem in a single sentence. It listens longer, questions assumptions, and finds patterns that others miss. Where it excels:

Long-form narratives
Executive speeches
Brand storytelling that needs tone and consistency

It’s less about throwing spaghetti at the wall and more about reducing the sauce.

The pairing

Used together, the effect is powerful. 4o opens the aperture wide; 5 brings it into focus. It’s the same creative dance that happens in any good team where one voice pushes for momentum, another for meaning. AI doesn’t replace that rhythm, it just gives you two new partners to rehearse with.

For creative or comms leaders, a few experiments worth trying:

Run a campaign brainstorm in 4o, then refine the shortlist in 5. What shifts?
Draft a leadership speech in 5, then ask 4o to sharpen the hooks. Do they align?
Brief 5 over several days on a long project and see how well it holds the thread compared with 4o.

The temptation is always to ask, “Which one is better?” The reality is more interesting: they’re different temperaments. What matters now is not just using GenAI, but curating how you use it, choosing the right collaborator for the moment, not the myth of a single, perfect machine.