爱剪AIcut

Screenshots

About

AICut edits video without a timeline. Tap + and choose what you are making: Talking head Pick a clip of yourself talking. An AI director listens, looks at a few frames, and hands you a plan: who it is for, which line opens, how it flows, what gets cut. Say yes, and an AI editor makes the cut: stumbles removed, your strongest line moved to the front, captions that highlight word by word, an animated opener, keywords that pop. A five-minute clip usually goes from picking the video to a finished cut in Photos in two or three minutes. It can also turn one long video into several short ones. Travel vlog Bring up to 20 clips. The director cuts each into shots, looks at the frames, listens to any narration, and first offers two or three ways to cut this footage; pick one and it plans and edits. Silent shots are fine, with an optional on-screen caption per shot; built-in royalty-free music ducks under speech. Create with AI No footage at all. Write what you want: the topic, the audience, the length. The director writes a storyboard for you to review; tap "Make it" and the app generates each scene's picture and voice-over, then assembles a vertical short. Pictures come from Volcengine Ark video generation, Ark image generation with a slow push-in, or a free typographic card, depending on the account you connect; voice-over uses Doubao speech. Want it different? Just say so: "Shorter, keep it under 45 seconds" "Switch the captions to the dark bar style" "Drop the third part" "Change the opener" · "Make scene two a night shot" Start right away 200,000 tokens of trial usage are built in. When it runs out, add your own API key under Models: DeepSeek, MiniMax, Volcengine Ark, or any endpoint using the standard chat API. Your video stays on your device Transcription, shot detection and rendering run on the device. Only the transcript and frame screenshots go to the model service; the video itself is never uploaded. In Create mode, each scene's picture description and narration text go to the generation service. Your API key lives only in the device Keychain. No account, no ads, no tracking. Chinese and English interface; edits videos in either language.Read more

AICut edits video without a timeline. Tap + and choose what you are making: Talking head Pick a clip of yourself talking. An AI director listens, looks at a few frames, and hands you a plan: who it is for, which line opens, how it flows, what gets cut. Say yes, and an AI editor makes the cut: stumbles removed, your strongest line moved to the front, captions that highlight word by word, an animated opener, keywords that pop. A five-minute clip usually goes from picking the video to a finished cut in Photos in two or three minutes. It can also turn one long video into several short ones. Travel vlog Bring up to 20 clips. The director cuts each into shots, looks at the frames, listens to any narration, and first offers two or three ways to cut this footage; pick one and it plans and edits. Silent shots are fine, with an optional on-screen caption per shot; built-in royalty-free music ducks under speech. Create with AI No footage at all. Write what you want: the topic, the audience, the length. The director writes a storyboard for you to review; tap "Make it" and the app generates each scene's picture and voice-over, then assembles a vertical short. Pictures come from Volcengine Ark video generation, Ark image generation with a slow push-in, or a free typographic card, depending on the account you connect; voice-over uses Doubao speech. Want it different? Just say so: "Shorter, keep it under 45 seconds" "Switch the captions to the dark bar style" "Drop the third part" "Change the opener" · "Make scene two a night shot" Start right away 200,000 tokens of trial usage are built in. When it runs out, add your own API key under Models: DeepSeek, MiniMax, Volcengine Ark, or any endpoint using the standard chat API. Your video stays on your device Transcription, shot detection and rendering run on the device. Only the transcript and frame screenshots go to the model service; the video itself is never uploaded. In Create mode, each scene's picture description and narration text go to the generation service. Your API key lives only in the device Keychain. No account, no ads, no tracking. Chinese and English interface; edits videos in either language.