The short version: We ran the same 12 ad briefs through GPT Image 2, Midjourney v7, and Flux 2 in 2026. GPT Image 2 won on character consistency, batch control, and prompt adherence. Flux 2 was the strongest pure-photorealism head-to-head. Midjourney produced the most cinematic stills but the most variance per generation, which kills batch work. If you are running performance creative at volume, GPT Image 2 is the default and Flux 2 is the backup.
Six months ago, most of the industry, including us, recommended Midjourney as the default for any performance creative work that needed photoreal output. The hero stills out of MJ v6 and v7 still look better than anything else when you are picking one image. But the ground has shifted in 2026, and Midjourney is no longer our default. This is not about anyone being wrong. It is about how this category actually works: image-model leadership flips every six months, and the only way to give an honest answer is to keep running the same test against whatever is current.
That is what we do. The moment a new image model ships, we test it against our current incumbents on the same brief set, so what we recommend to clients is based on what is actually leading today, not what was leading last year. Performance creative asks: can you produce 20 variants of the same character, in 8 different lighting conditions, on a Tuesday morning, without watching a Discord server. That question has a very different answer than “which model makes the prettiest single image.”
This post is the comparison we ran across 12 internal briefs in 2026. It is not exhaustive. It is honest about where each tool wins and loses when the job is producing ad creative at volume.
What does “best for performance creative” actually mean?
We use the same five-question rubric every time we test a new model. We call it the Five-Lens Image Test™. Most public benchmarks only look through the first lens.
- Photorealism. Does the face look real on a 6.7-inch phone screen at full brightness?
- Character consistency. Can you make the same person show up in 12 different scenes and still look like the same person?
- Prompt adherence. When you ask for “Korean woman, 28, in a kitchen at golden hour, holding a phone,” do you get exactly that, or do you get whatever the model felt like that day?
- Batch control. Can you script the prompt? Version-control it? Re-roll one variant without re-running the whole set?
- Cost per usable variant. Not cost per generation. Cost per image you actually ship.
Twitter benchmarks usually only score #1, and even structured leaderboards like the Artificial Analysis image-model arena lean heavily on best-of-set photorealism. That is a hero-image question. Performance creative is a workflow question, and the workflow is what we score.
The 12-brief test
We picked 12 briefs across the kinds of work we ship for clients: train commuter, bedroom confessional, golden-hour window selfie, group chat reaction, desk workspace, two-creator split, kitchen scene, and a few more. We ran each brief through all three models, tweaking the prompt format to fit each one. Then we scored on the five questions above using a blind 1-to-5 rubric.
Here is the summary, with detailed notes per model below.
| What we tested | GPT Image 2 | Midjourney v7 | Flux 2 |
|---|---|---|---|
| Plans the image before drawing it (reasoning) | Yes | No | No |
| Photorealism (average frame across 12 briefs) | 4.5/5 | 3.5/5 | 4.5/5 |
| Character consistency across 12 variants | 4.5/5 | 2.5/5 | 4/5 |
| Prompt adherence | 5/5 | 3/5 | 4/5 |
| Batch control / scriptability | 5/5 | 1/5 | 4/5 |
| Cost per usable variant | $0.20 to $0.50 | $0.40 to $1.20 | $0.10 to $0.30 |
| Best fit for performance creative | Default | Hero stills only | Cost-sensitive backup |
The “cost per usable variant” line is the one most people get wrong. The right question is not what one generation costs. It is how many generations it takes to land on an image you actually want to ship, times the per-generation cost.
GPT Image 2
GPT Image 2 is OpenAI’s newest image model, released on April 21, 2026. We use it through the API and the Codex CLI, and it is our default.
Here is the part most reviews skip, and the reason we switched. GPT Image 2 is the first image model that actually thinks about the picture before it draws it. Midjourney and Flux are diffusion models. You hand them a prompt, they turn that prompt into a vector, and they translate the vector to pixels in one shot. There is no planning step. They do not check their work.
GPT Image 2 does. According to OpenAI’s own prompting guide, the model runs a reasoning pass before generation. It figures out what the scene needs, where each element goes, and what the brief is really asking for. Then it generates. Then it checks the output against the brief and re-renders if it missed something.
That sounds abstract until you see it in practice. Ask Midjourney for “two people in a kitchen, the woman holding a coffee mug.” Half the time the man ends up with the mug, because the diffusion process never asked itself who has it. Ask GPT Image 2 the same thing and the answer is right almost every time, because the planning pass settled the question before the pixels existed.
That reasoning is what powers everything else. Prompt adherence stops being a coin flip. Character consistency holds across 12 scenes if you describe the character once and reuse the description. Multi-element layouts, on-image text, hand positions, eye-line direction. All the things diffusion models used to garble, GPT Image 2 gets right first try most of the time.
What GPT Image 2 loses on: pure cinematic feel. Midjourney still produces more “wow” stills if you are picking one image for a billboard. For performance creative, that does not matter. The algorithm and the audience care about consistency and specificity, not magazine cover energy.
Midjourney v7
Midjourney v7 produces the best individual hero shots in this comparison, full stop. If you are making one image and you want it to look like a film still, Midjourney is still the answer.
For performance creative, it is not. The two structural problems are batch control and prompt adherence.
Batch control is the bigger issue. Midjourney runs through Discord. There is no first-party API. Third-party wrappers exist but are unofficial, rate-limited, and break when Midjourney updates. We tried scripting MJ for a 12-variant sprint and burned a half day on plumbing. Same sprint in GPT Image 2 took 20 minutes.
Prompt adherence is the smaller issue but matters at scale. Midjourney has its own aesthetic point of view, which is great when that aesthetic matches your brand and brutal when it doesn’t. Asking for “subtle, candid, slightly imperfect lighting” returns something cinematic and intentional anyway.
Where Midjourney still wins: hero brand stills, key art, anything that gets used once and seen many times. Different job.
Flux 2
Flux 2 is the open-weights wildcard. It runs on fal.ai, Replicate, or self-hosted, which makes it the cheapest of the three per generation by a wide margin. According to fal.ai’s published pricing, Flux variants run a small fraction of a dollar per image at standard resolution, and on Replicate’s Flux 1.1 Pro listing the per-image cost is similarly low. Both run well below GPT Image 2’s API rate at the same resolution.
Photorealism on Flux 2 is genuinely close to GPT Image 2 across most face types. Character consistency is slightly weaker. Prompt adherence is good but not as strong as GPT Image 2. Batch control is strong because the API is open and well-documented.
Where Flux 2 wins: cost-sensitive sprints where you want to generate 100 variants cheaply and curate down to 20. The low cost per image lets you over-generate and pick.
Where Flux 2 loses: face consistency across long sprints, and edge cases in non-Western character types where the model’s training data is thinner.
Our pick and how we use them
For most performance creative work, we reach for GPT Image 2 through Codex. The reasoning pass plus the character consistency is exactly the tradeoff paid social needs.
For high-volume, cost-sensitive sprints, we use Flux 2 through fal.ai. The per-image cost lets us over-generate and pick the best ones.
We do not use Midjourney for performance creative anymore. We still love it for hero brand work and pitch decks, where one beautiful still does the whole job.
If you want the one-glance answer, this is the cheat sheet we share with clients:
| If you are doing this | Use this | Why |
|---|---|---|
| 8+ variants of the same person across different scenes | GPT Image 2 | Reasoning pass keeps the character consistent |
| 100+ variants where unit cost matters most | Flux 2 via fal.ai | Cheapest per usable image at scale |
| One cinematic hero still for a deck or billboard | Midjourney v7 | Best single-image craft |
| Anything with on-image text, signage, or packaging copy | GPT Image 2 | 95%+ text accuracy, no other model is close |
When this comparison will be wrong
One caveat. Image model leadership flips every six months. A year ago Midjourney was the obvious default. A year from now Flux 3 or some open-weights model we have not heard of yet could be the answer. The Five-Lens Image Test™ outlasts any specific model. The specific picks will not.
If you are setting up a performance creative pipeline in 2026, GPT Image 2 is the default. If you are reading this in 2027, run the Five-Lens Image Test™ against whatever models are leading then, and pick again. We will too.
Where this fits in the larger workflow
This is the image-generation step of a three-step pipeline. The other two are creative direction (a human creative director paired with a Claude Code agent) and animation (Seedance 2.0 or Kling 3.0, batched through Lovart). We covered the full workflow in our AI performance creative case study, including the campaign that dropped CPA 50% on TikTok.
If you want to talk through what an AI performance creative pipeline could look like for your business, that is exactly what our service is built around.
Like this? Get the next one.
Short emails. New posts as they ship.
