Image Tools · AI Image · Slides · Nano Banana · GPT-Image-2 · Seedream 5.0 Pro
Use AI Image for individual images and region edits with version history. Use Slides for structured decks with slide content and narration. Finished images can be passed to Image to Video or Reference to Video when motion is needed.
Nano Banana, GPT-Image-2 and Seedream 5.0 Pro sit in the same editor, so the same prompt can be run through any of them and the results stay side by side in version history. Start here for covers, thumbnails, and anything built from a written description or a reference photo.
When one frame cannot carry the point, Slides turns a topic, a file, or a link into a deck where every page has real content and a spoken script. Leave the page count on Auto or set it yourself, pick a visual style, and switch on Visual Only when you want the pages without a voice track.
Do not regenerate the whole frame. Select the area that is off, describe what it should be instead, and only that region is redone. The edit is saved as a new version, so the one you started from is still there to compare against.
AI Image stops at the still. Download the finished picture and upload it in Image to Video, where it becomes the first frame, or in Reference to Video, where references fix the person, product and style while the prompt decides the camera motion. AI Video is the head page if you have not settled on a starting point.
AI Image makes one picture at a time from a prompt or reference photos, and produces no speech and no video. Slides makes a structured deck from a topic, a file, or a link, where each slide has its own content and a narration script, and it can also export a narrated video.
Nano Banana, GPT-Image-2 and Seedream 5.0 Pro, all in the same editor and switchable on the same prompt. You can add up to five reference images in JPG, PNG or WebP, choose from eight aspect ratios between 1:1 and 21:9 and three quality levels, and generate one, two or four images per run.
Yes. Select the region, describe the change, and only that area is regenerated. The edit is kept as a new version rather than replacing the previous one, so you can compare versions, return to any of them, and download the original file.
A deck from a topic or prompt, from files you upload (PDF, Word, EPUB, Markdown, TXT), or from a link such as an article or a YouTube video. Length is Auto or a fixed 1 to 20 slides, the visual style is yours to pick, and narration is read in English, Chinese Mandarin or Japanese by a voice from the library or your own cloned voice. PPTX and PDF export always; with narration on, also the narrated video, the audio track and SRT subtitles.
Download it and upload it in one of the video tools. Image to Video takes a still as the first frame (JPG or PNG, up to 20 MB per image), with a last frame optional for a transition, and clips run up to 15 seconds. Reference to Video takes up to nine reference pictures, which decide the person, product and style while the prompt decides the camera motion — references do not set the opening frame. The credit cost is shown next to the generate button before you confirm.
They can be used in personal and commercial work. Check the current terms for what applies to your plan.