I built a ChatGPT Dot that automatically clips content for you. Send it a URL and wake up to clips already posted on X, TikTok, YouTube, Reels, and LinkedIn. Every clip scored for virality, reframed with speaker tracking, and captioned in your brand template. Comment "DOT" and I'll send it over + FREE credits (must be following) 👇

We just brought @OpusClip to @ChatGPT! You can now clip videos, repurpose content, and schedule posts in the same conversation. But the interesting part is that you don’t need to know exactly what you want to make before you start. You can drop in a video and ask: “What’s worth sharing from this?” “Which moments have the strongest hooks?” “Could this work as an X thread?” Then take the ideas you like and actually do something with them. Create captioned clips, write copy for different platforms, outline a carousel, or plan and schedule your posts. You can be specific, too: “Find the part where we talk about pricing.” “Make these titles less clickbaity.” “Which of these clips feel repetitive?” A lot of people already have useful things to share sitting in podcasts, interviews, webinars, and recordings. Turning that material into something people actually see is where they get stuck. You still review the work and decide what goes live. But you can go from exploring an idea to getting it published without starting over in a different tool at every step. Just @OpusClip in ChatGPT!

I built a Grok Bot that automatically clips content for you. Send it a URL and wake up to clips already posted on X, TikTok, YouTube, Reels, and LinkedIn. Every clip scored for virality, reframed with speaker tracking, captioned in your brand template, titled to stop the scroll. Comment "BOT" and I'll send it over + FREE credits (must be following) 👇

We might've solved video repurposing > Reformatting today means black bars or cropping > Either way, your footage gets worse just to fit a new platform > Changing aspect ratio should adapt your video, not destroy it That's why we released AI Expand: reformat any video by generating outward, not cropping in Your original frame stays fully intact. The scene extends beyond the edges and blends with what you actually shot. Nothing cropped, nothing padded. One clip, ready to post anywhere.

We just shipped Reimagine! Most AI tools restyle by regenerating from your prompt. You get a different video (usually a worse one) because what you liked was never in the prompt. Reimagine fixes that: > keeps same scenes, order, & timing > applies style across all shots

Question I had: how much does it actually need from me to make something decent? So I gave it nothing. I just prompted “ad for a fake meal kit service.” Agent Opus made a fake logo, product, and name and generated this with Seedance 2.5. Here's the result 👇

This entire video was generated in one prompt in @AgentOpusAI By far the easiest way to make explainers, ads, skits, and other AI videos without needing to prompt engineer. Comment "PROJECT" and I'll send you the project for free :)

TaxAct is publishing YouTube explainers without writing a single new script. Every video is a blog post they already published years ago. TaxAct is owned by Taxwell, which also owns Drake Software and 1040.com. Between them: 90,000+ professional tax preparers and 3 million+ DIY filers a year. Go look at the videos. Then go look at the blog. > Same topics > Same order > Same structure The archive was the script library the whole time. Anyone can run this: 1. Open your oldest evergreen blog post 2. Hand it to @AgentOpusAI 3. Generate the explainers 4. Publish 5. Tomorrow, next post And they're embedding the videos back into the original blog posts. It's distribution. The blog post ranks in Google. The YouTube video ranks on YouTube. The embed turns readers you already have into views on a surface your blog can't reach. Nobody is writing new content here. They're re-rendering content they already paid for into the format the audience actually watches. If your company has a blog older than two years, you are sitting on a video catalog and treating it like a text archive.

100% of this video was generated inside Claude. Built an internal tool for building launch videos with code.