OpenAI · 2024 · 12 · 09 · Tool · ~2 min read

OpenAI made Sora generally available

Sora — OpenAI's text-to-video AI — left research preview and became available to ChatGPT Plus and Pro subscribers. Up to 20-second 1080p clips, with editing tools and a remix feature.

What's actually new

  • Public access through chatgpt.com — no waitlist for the first time. Plus tier got limited credits, Pro tier got plenty.
  • Storyboard editor for chaining shots together, plus a 'remix' button to nudge an existing clip into a variation.
  • Higher quality than the original February 2024 preview — longer clips, sharper picture.

If you want more

Worth knowing~30s
  • The showcase reel hides the failure rate. Real prompts often produce melting hands, drifting scenes, or objects that morph mid-clip.
  • The Pro tier costs US$200 a month. You need it for unlimited generation. Plus tier hits limits quickly.
  • Banned in EU and UK at launch — 'available' wasn't actually global on day one.
Who should care~20s

Creators looking at AI video for quick prototypes. Teachers thinking about how to talk to students about AI-made media. Journalists covering deepfakes. Anyone whose job involves trusting moving images on the internet.

What to do about it~20s

Try a few prompts to calibrate your sense of what's plausible. Then assume any short clip you see online could be Sora-made — especially if it looks slightly too smooth. Watermarks help. They're not reliable.

Honest take~45s

The general release of Sora was the moment AI video crossed from 'impressive demo' to 'thing your cousin can use'. The bigger fallout isn't for the creative industry — most TV-quality work still wants real cameras. It's for trust. For social-feed video, news clips, and scams, the floor just dropped. Provenance — proof of where a clip actually came from — is now urgent infrastructure, not a nice-to-have.

Other recent tool updates

Sources

Last verified · 2026 · 05 · 05 · Found a fact wrong? corrections@aguidetocloud.com