Press & Data
Data and comparisons for journalists
PromptFrenzy runs the same prompt through every major AI image and text model and publishes the results — head-to-head, with live pricing and a release timeline. If you're writing about AI models, we're a source: we have the data, and we're happy to run a specific comparison for your story.
Story angles we can speak to
- →Which AI model is actually best — and cheapest — for a given task
- →How the frontier models compare head-to-head (GPT vs Gemini vs Claude)
- →What AI image and video generation really costs in practice
- →Whether a newly-released model is genuinely better than the last
- →Prompt engineering and AI image generation for everyday users
Latest report
Physics Reasoning: implicit optics in AI image models
Google Nano Banana 2 was the only tested model to apply the implied optics reliably, passing five of six runs (83.3%); every other model passed at most one of six. Read the 42-run benchmark →
The data, live
Everything below updates continuously — link straight to it, or cite the figures with the line under “How to cite us.”
Head-to-head comparisons
The same brief through every model, side by side.
Benchmarks
Curated prompts run across the image models, with outputs.
Latest research dataset
Download every Physics Reasoning result as citation-ready CSV.
Model directory & pricing
Every model — what it is, what it costs, real examples.
Release timeline
When each major AI model launched, newest first.
Most recently added models
How to cite us
Please attribute figures and comparisons as:
Source: PromptFrenzy — promptfrenzy.com
A link to the specific page (e.g. the comparison or the release timeline) is appreciated. Need a custom comparison, a high-res chart, or a quote? Email maria@promptfrenzy.com.
Writing about AI models?
Tell us the comparison you need and we'll run it — same prompt, every model, with the receipts.
Get in touch →