Hello, everyone!
It's not a title party. After having AI editing, I really rarely open editing.
Friends who make videos should be very familiar with cutting and screening. In the past, when cutting an oral video, I would usually use "Smart Preliminary Cutting" and "Smart Clipping" first, then check word by word whether there was a mistake, and finally use "Smart Packaging" to process the video. Media Processing Service. The whole process may not seem complicated, but it is still time-consuming to actually do it.
Now that we have entered the AI era, the editing method is completely different. In the past, you had to do it little by little, but now many things can be done directly through dialogue:Just tell the AI how to cut, add what effects, and do subtitles.
What I want to introduce to you today is a set that I like so far AI Clipping 0 Cost Plan: Google Antigravity + ChatCut.
This solution can realize intelligent cutting and broadcasting, automatic adding of subtitles, BGM, dynamic effects and other operations. For example, you can directly tell it: "Help me cut this oral video, remove modal particles and stutters, and make the overall smoother." It will automatically process it according to your request.
What is ChatCut?
ChatCut is an AI editing software that only appeared this year. It is generally similar to cutting and screening, but its biggest features areSupport for controlling editing through Agent dialogues。
It comes with its own Agent, but using the built-in Agent requires points. In addition, it also supports accessing external Agents, including Codex, Claude and other Agents.
That is, if we use our own Agent to control Cut,The editing function of the software itself can be basically used for free.
So, combining ChatCut and Antigravity can achieve a near-zero cost AI conversational editing solution.
How to connect to Antigravity?
After downloading, installing ChatCut and logging in, the entire interface is still very simple. The core is uploading video materials.
After entering the software, find the top **"Connect to Any Agent"**. After clicking in, you will see a series of connection instructions. Copy this string of instructions and paste them into your Agent to connect to ChatCut through MCP.
ChatCut also supports direct access to Codex and Claude, but these two require their own corresponding members, and if a large number of videos are edited, Codex's Token consumption is quite obvious.
So I recommend using Google Antigravity + Gemini Flash This combination.
After copying the connection instructions provided by ChatCut to Antigravity, the connection is successful, which is equivalent to moving the conversation control function of ChatCut to Antigravity.
The way I use it is usuallyPut Antigravity on the left and ChatCut on the right。
Upload the material to ChatCut first, then go back to Antigravity and tell it directly how you want to cut it.
For example, for oral video, you can directly enter:
Help me edit this oral video and remove modal particles, stuttering and non-smooth parts to make the overall rhythm more natural.
Then it will automatically start working.
Of courseThe more specific you describe your requirements, the better the end result will usually be.
If you don't know how to describe it, there is also a very easy method:Directly upload a finished video you like and let the AI refer to its editing style.
For example, if you like a blogger's video rhythm, subtitles, titles, transitions, and animation, you can ask the Agent to refer to the video and then process your material in a similar style.
What is the effect of AI cutting?
I actually used it myself, and the effect was quite good.
I have also helped others cut oral videos before, especially videos recorded by some bosses. There are often problems such as stuck, repeated, and many modal particles when speaking. In the past, after using clipping and reflection to make intelligent preliminary cuts, it still required manual careful inspection, sometimes even to correct them word by word to see which areas should be deleted and which areas should be retained.
Now that it is handed over to AI, this part of the work can be basically completed automatically.
It will not only get rid of some obvious jams and modal particles, but also judge what content should be retained based on the context, making the overall oral broadcast smoother. In some places, zoom-in, subtitles, animation, etc. will be automatically coordinated to make the video not appear particularly stiff.
Most importantly,AI editing does not end after being generated in one go.
If you feel dissatisfied with anything, just tell it directly: "The pace is too fast here","Don't zoom in on this place","Change the subtitles to a different style","Cut this paragraph again", and let it continue to be revised.
You can let it be optimized repeatedly, do other things yourself, and come back and check it out later. If you are not satisfied, then issue another instruction to continue modifying.
The biggest difference between this and sitting in front of the computer and cutting it bit by bit is:People have changed from "editing in person" to "reviewing + giving instructions".
Why do I recommend Antigravity?
I actually used Codex and Claude to control ChatCut at the beginning, but if there are many videos, the Token consumption is still quite obvious.
Now switching to Google Antigravity, coupled with Gemini Flash, the biggest advantage isFast speed and large usage。
So when cutting videos, you don't have to worry too much about not enough Tokens. Especially when making streaming videos, when you need to process many pieces of material in a day, this combination will be much more convenient.
If you have multiple Google accounts, it will be more flexible to use.
Of course, the real core of this solution is not that we have to use Antigravity.
As long as your Agent can connect to ChatCut normally, a similar effect can be achieved.
So if you can't use Google Antigravity, you can also try using other domestic Agents. The core steps are all the same:Copy ChatCut's connection instructions → paste it into Agent → connect → upload material → issue editing instructions in natural language.
The whole process basically follows:
Upload the material → Tell the AI how to cut it → AI automatically edits → Manual review → If you are not satisfied, continue to let the AI modify it → Export.
This is my favorite AI editing method at the moment.
We used to spend a lot of time sitting there cutting videos, but now we only need to be responsible for "telling the AI what to do" and "finally checking the effect."
The most important thing is that this plan focuses on a relatively complete AI video editing process that is almost zero-cost and can be completed without cutting and screening members.