Changelog
See the latest TapNow features you can use right away.
v2.13.3
Seedance 2.5 now supports 1080p
When you use Seedance 2.5 to generate a video on the canvas, open Resolution and select 1080p. Frames, Omni Reference, and Video Edit all support this option. New tasks still start at 720p.
See the video generation tutorial →v2.13.1
Answer several Agent questions at once
When Agent needs more than one piece of information, its questions appear above the input. You can answer up to four questions one by one and submit them together. The answer summary remains available in the conversation.
See the Agent conversation tutorial →Choose the quality for GPT Image 2 Outpainting
Select Outpainting on an image node and use GPT Image 2 to choose Low, Medium, or High quality. New tasks use Low by default.
See the image editing tutorial →v2.12.29
Preview 3D content before opening a studio
Open a 3D asset, panorama, or 3D scene to rotate the object or look around first. Select Open Viewfinder, then Capture, to add the photo directly to the canvas. When you want to continue building the scene, select Use in 3D Studio.
See the 3D tutorial →Switch among three models when editing images
Select an image node, then choose GPT Image 2, Banana Pro, or Seedream 5.0 Pro for Redraw, Outpainting, and Erase. The result appears as a connected new node to the right of the source image, while the source remains unchanged.
See the image editing tutorial →Use Library elements as MiniMax H3 references
Choose MiniMax H3 in Reference mode, then select Add element to use a visual reference from the Library. This option is not available in Multi-Shot mode. You can use up to nine images, three videos, and three audio clips at a time. Audio must be used with at least one image or video.
See how to use elements →v2.12.28
Ask Agent to explain how a TapTV project was made
Open a TapTV project preview on desktop and select Explain How It’s Made. TapNow clones the project, opens an editable canvas, and asks Agent to explain the creative approach through its nodes and connections. The original public project remains unchanged. You need to sign in, the action uses about 40 Tapies, and it is not currently available on mobile.
See the TapTV tutorial →v2.12.19
Extend a video and retake individual segments
Select a video node and use Continue video · Beta to add 4–30 seconds before or after it. You can also use Video Retake to set a new view, shot scale, or camera move for each time segment. Both tools are in Beta. The result appears as a connected new node, while the source video stays on the canvas.
See the video editing tutorial →v2.12.18
Generate with Seedance 2.5 on the canvas or in Agent
Choose Seedance 2.5 in a video node or an Agent confirmation card, then set the mode, references, a 4–30 second duration, and output settings to generate the result on the canvas. Reference mode accepts up to 30 images, 10 videos, and 10 audio clips. Model availability depends on the options shown for your account.
See the Seedance 2.5 tutorial →MiniMax H3 adds 768P, four-second video, and nine reference images
Choose MiniMax H3 in a video node to set Resolution to 768P or 2K, choose a duration from 4 to 15 seconds, and add up to nine images in Reference mode. New tasks still start at 2K and five seconds. Check the updated Tapies estimate before generating.
See the MiniMax H3 tutorial →v2.12.13
Generate audio with Seed audio 1.0 and references
Choose Seed audio 1.0 in an audio node, enter a prompt, and add up to three audio references before setting the format, subtitles, and advanced controls. The result appears as a playable waveform. When subtitles are enabled, a connected audio-subtitle text node is also created. Image references are not currently supported.
See the audio generation tutorial →v2.12.10
Search History and return multiple results to the canvas
Open History from the left side of the canvas to search generated images, videos, audio, and 3D content, or switch between All projects and This project. Choose Select to apply or download several items at once. History stores generated results, not a snapshot of the entire canvas.
See the History tutorial →v2.12.9
Generate video with MiniMax H3
Choose MiniMax H3 in a video node to generate video with native audio through Frames or Reference. At launch, it supported 2K, 5–15 seconds, and up to five reference images. See the later version above for its expanded settings.
See the MiniMax H3 tutorial →v2.12.5
Reuse people, products, and characters with Elements
Open Library and Elements from the left side of the canvas to create an element from canvas content or a local file, then save it to your Private or Team space. In a supported model’s Reference mode, use Add element or @ to reuse the same person, product, or character.