NightCafe has a free side-by-side comparison of models. Open it at creator.nightcafe.studio/evals. That address opens the text-to-image grid. Use the tabs at the top of the page to switch.
There are four public grids:
You can also find it from:
- The What's New popup and the What's New page. AI Model Comparison covers the image grids. Video Model Comparison (10 November 2026) covers the video grids.
- Your profile menu — AI Model Comparison (under My Collections). This opens the text-to-image grid; use the tabs for the others.
- The site footer — NightCafe column, under Blog. Same text-to-image link.
- Studio's image model picker — a discovery note that says “Not sure what model to choose?” with a Compare Models button (opens in a new tab)
- Studio's video model panel — a Compare Models button next to All Video Models and Usage Guide. It opens the text-to-video comparison.
- The All models picker while you are in video mode — the same discovery note, linking to text-to-video
- The Animate window — the model description links to the image-to-video comparison, and its All models view has the same Compare Models button
- Prompt to Edit — the edit-options model picker and the “Choose Prompt to Edit Model” modal link to the Image Editing comparison
What you see
Each row is the same prompt. On Image Editing, the row also uses the same start image or set of start images. On Image to Video, the row can use the same starting image, the same reference images, or a start frame and a separate end frame.
Each column is a model. Some models appear more than once when they have more than one quality setting. For a finished image you can see the credit cost, how long that generation took, the output size, and whether the model is PRO. For a finished clip you can see the credit cost, how long that generation took, the clip length, and whether the model is PRO. The model name also appears under each result.
- Click a thumbnail to view an image larger, or to play a clip.
- Try in Studio opens that model in Studio.
- On Image Editing, models that only accept one start image are skipped on cases that use more than one input.
- On video, a model that cannot run that kind of input is left blank. Clip length is shown on each result. A model uses the closest length it supports, so two columns in the same row are not always the same number of seconds.
How to browse
- Use the Models control to show or hide models, drag to reorder them, or sort by Default, Recently added, Cheapest, Most expensive, or Fastest.
- On Text to Image, use the category menu to focus on Art & Illustration, Tricky Prompts, Typography & Text, Portraits & People, Design & Products, Photorealism, or Anime & Manga.
- On Image Editing, categories include everyday edits and a Style Transfer set — useful if you used to restyle photos with the original Style Transfer algorithm.
- On Text to Video, categories include Art & Animations, Realism, and Tricky Prompts.
- On Image to Video, those groups can appear, plus Reference to Video (a clip from reference images rather than a first frame) and Start & End Frame (a clip that moves from a first frame to a last frame). A category only shows when there are examples in it.
- You can pin the prompt column so it stays visible while you scroll sideways.
- You can change thumbnail size (small, medium, or large).
- Long prompts can be expanded with Show full prompt.
Credits and scoring
Looking at the comparison does not spend credits. It is a visual comparison of NightCafe's own benchmark prompts, not a third-party leaderboard and not a ranked score. Credit costs shown on each result can change over time — check Studio for the current price before you generate.
For written recommendations (including value picks such as Muse Image, GPT Image 2.5 Low, Flux 3 Klein, and Seedream 5.0 Flash), see the NightCafe AI Models Guide. For how to make a video, see Creating Videos on NightCafe.
FAQs
Does opening the comparison use credits?
No. Browsing the grid is free. You only spend credits if you generate something in Studio.
Can I try the same prompt myself?
Use Try in Studio on a result to open that model. The comparison prompts are NightCafe's benchmark set — you can write your own prompt in Studio.
What's the difference between the tabs?
Text to Image starts from a prompt only. Image Editing starts from a prompt plus one or more existing images. Text to Video starts from a prompt and shows short clips. Image to Video starts from a prompt plus a starting image, reference images, or a start frame and end frame.
Is this the same as a live workshop?
No. The comparison page is always available. NightCafe also hosts occasional live Create With Me sessions — the latest replay is Make Something for the Season.