# Can a ChatGPT Dot make videos? Editing, Blender 3D, music videos and explainers: 43 real projects

Yes. A ChatGPT Dot (OpenAI Dots) makes video on its own cloud computer: Blender (preinstalled, no GPU), Remotion and Python run there, while generated footage, songs and voices come from your own Higgsfield, Suno, ElevenLabs or ComfyUI setup. 43 real projects, from a 26-second draft made in five minutes to a 2:51 music video built from a Suno song and old photos.

Can a ChatGPT Dot make videos? A short version first: what it does well, seven setup steps covering what its cloud computer has (Blender, no GPU), how your footage reaches it, connecting your own generative tools, briefs and first cuts, overnight renders and keeping publishing for yourself, where video work gets stuck and what worked, and one message to start with. Then quick answers, first-hand projects (edits, explainers, Blender 3D, generative pipelines, music videos) and more asks you can copy.

### Can a ChatGPT Dot edit or make videos?

Yes. Each Dot has its own Linux cloud computer and keeps working while your laptop is closed ([Introducing dots](https://openai.com/index/introducing-dots/)), so you can ask for an edit, hang up and come back to a file. Nusa Ardi W got four edits of the same footage, each in under an hour; SEOWOO had a keynote clip translated and subtitled in Korean; 이국환 had a 26-second draft five minutes after sending a link from the car; のえる's promo video arrived 17 minutes after he hung up.

### What runs on the Dot's computer, and what needs my own tools?

Blender, GIMP, Remotion and Python ran on the Dot's machine in these cases. Owners who inspected it found Blender preinstalled and no GPU, so KANA's palace came out as CPU-rendered stills cut into a 33-second tour (details on [Cloud computer](/use-cases/cloud-computer/)). Generated footage and voices came from the owner's own accounts: Higgsfield, Minimax H3 through ComfyUI, ElevenLabs, HeyGen, Fish Audio and Suno.

### Can it build 3D models in Blender?

Yes, usually as a first version you refine. Antony Budianto got a folding-phone model with an animated hinge as a ZIP with four preview renders; 齋藤和哉's Dot modelled his company's building from photos and a public floor guide, with estimated dimensions; 강민재 posted his car model as a progress report that still needs refining.

### How do I get my own footage to the Dot?

Not always through chat. 大東信仁's two-minute iPhone video failed with "file too large" (as far as he could tell, only files up to 20 MB went through), while Allison Johnson's TV clips were simply files on the computer she had connected through the ChatGPT desktop app, after granting a lot of permissions. For longer files, use a connected computer or [Google Drive](/integrations/google-drive/).

### Can I leave a long render running overnight?

Yes. 日沼諭史 asked for a 3D animation of the next day's weather around Tokyo Station and found it finished the next morning: about four and a half hours of gathering data, modelling and rendering in Blender, and exporting. Work you assign also keeps going after a call ends ([dots guide](https://learn.chatgpt.com/docs/dots)).

### Will it publish to YouTube on its own?

It can upload, so decide the last step yourself. YMinamiyama's voice drama went up as private, Kanman's Dot started with a 5-second unlisted test and then noticed the Google connection would expire after seven days, and Adem Vessell gave Morrow its own YouTube account; it now voices, edits and uploads its own videos to TikTok, YouTube and X. When the YouTube player never loaded, Loïc Frissard's Digi said so and asked for the transcript. To turn a finished video into posts and newsletters, see [Content pipeline](/workflows/content-pipeline/).

### Which plan do I need?

A Dot comes with Pro ($100, $200 or $500 tiers) outside the EEA, UK and Switzerland, with Business Premium in any supported region, and with Enterprise once an admin switches it on ([How to get ChatGPT Dots](/access/); in Europe, see [Dots in the EU and UK](/dots-eu-uk-availability/)). The full list of what a Dot can do is on the [features page](/features/). If the Dot goes quiet, start at [Dots not showing or not working](/dots-not-showing/).

Source: https://gptdot.net/use-cases/video-and-animation/