Run one prompt through two AI image models at once
Stop guessing which model suits the brief. Ideovex renders the same prompt on two of its 39+ image models in the same window, at the same size, and lets you call the winner.
How it works
Three steps, one window.
No project setup, no downloads, no switching tabs between two platforms to answer one question.
Pick two models
Any two image models in the library, free or premium, in either column.
Write one prompt
Both sides get the same words and the same frame size. Nothing is tuned behind your back.
Call the winner
Mark the side that won, download it, or send it straight into the studio to keep working.
Popular pairings
The comparisons people run first.
Every pairing below is live in the studio right now, with the real credit cost of running it.
Flux 2 Max
vsGPT Image 1.5
The two flagship stills. Flux 2 Max for texture and light, GPT Image 1.5 for prompts that read like instructions.
Nano Banana Pro
vsSeedream 5 Pro
Both premium, both fast. The split usually comes down to faces and skin against graphic, high-contrast composition.
GPT Image 1.5
vsSeedream 4.5
The typography question. GPT Image 1.5 reads a brief like instructions, which is what you want when the image has words in it.
Flux 2 Dev
vsSeedream 5 Lite
The 1-credit bracket, where you are deciding what a cheap draft is allowed to look like.
GPT Image 2
vsNano Banana Pro
Two free models against each other, so the comparison itself costs nothing.
Midjourney V7
vsNano Banana 2
House style against literal reading — Midjourney interprets, Nano Banana 2 obeys.
What you get
A fair test, and both files.
Identical brief
One prompt, one aspect ratio, two renders — the only variable left is the model.
A verdict you keep
Your winner is recorded against the run, so the choice survives the session.
Both files
Download either side at full resolution. Losing a comparison does not bin the image.
Carry on working
Send the winner into the image studio with the prompt intact and iterate from there.
39 image models to choose between
4 of them run free on the experimental pool, and paid stills start at 1 credit — so a comparison can cost nothing at all or 500 credits at the flagship end.
FAQ
Questions before you run one.
Each side is charged at its own rate, so a comparison costs the two models added together. Two 1-credit models is 2 credits; a 20-credit flagship against a 1-credit draft model is 21. Free experimental models cost nothing, so a free-against-free comparison is free. The exact number is on the button before you run it.
Yes. The prompt and the frame size are sent to both models unchanged — no per-model rewriting, no hidden style presets. That is the whole point: the output difference you see is the model, not the brief.
Yes. The free experimental pool is in the same picker as everything else, so you can run two free models against each other and spend nothing. Free models share capacity, so they are slower and can ignore some settings.
Yes, because a render costs real compute. A free account is enough — it comes with daily credits, no card, and reaches the full model library including the premium models.
Side-by-side is image-only today. Video and music models run in their own studios, where the controls that matter there — clip length, start frame, resolution — have room to exist.
Credits are reserved when a run starts and settled when it ends, so a failed side is refunded automatically. The side that did land stays on screen.
Settle it with a render, not a review
Free account, daily credits, no card. Your first comparison can be running in two minutes.