SORA-2 dropped last week—and it’s already the hottest thing since VEO-3.
It’s still brand new, so deep dives on how to create top-quality videos are rare.
But we’ve gathered some solid tips from multiple sources—and organized them for you here.
Anyway, here are the tips:
#1 - High Quality Onboarding Video
Unlike VEO-3, SORA-2 doesn’t let you upload a photo or video of someone as a starting point for your creation.
Instead, SORA uses a feature called Cameos—meaning you and others can appear in the same video.
You’ll need to record a short video of yourself (including you speaking) during setup so SORA can generate videos that look and sound like you.
Make sure your video is recorded:
In good lighting
Against a contrasting background (ideally a green screen)
In complete silence (so only your voice is captured)
Also, the Cameo feature is optional—you can turn it off if you don’t want others using your face or voice.
With that said, you can only create cameos with others if you keep the feature turned on —that’s how people have been making those Sam Altman-style videos.
Either way, it’s a really fun feature. If you and your friends record solid onboarding videos, you can easily generate high-resolution duo clips together.
#2 - Cartoon Videos Work Well
Styles like Minecraft, anime, South Park, or any other recognizable cartoon format tend to produce the most accurate outputs.
Regular human-style videos can be inconsistent—some turn out crisp and realistic, while others look completely off.
So even if you don’t plan to make an Avatar or SpongeBob-themed video, know that established cartoon styles usually deliver higher-quality results.
#3 - Understand Its Limitations
Unlike VEO-3 or Kling, Sora has a relatively small prompt limit—capped at just 2,500 characters.
This is mainly because Sora functions more like a social app than a production suite.
With VEO-3, which connects to Google’s Flow software, you can create cinematic, storyboarded videos using a full-scale editor.
Sora, on the other hand, isn’t designed for that level of complexity.
OpenAI designed Sora so anyone could get value from it without needing complex prompts. The goal was to let users create high-quality videos using natural language alone.
So if you’re an advanced prompt writer, you’ll likely need to simplify your style.
Also, note that Sora videos are capped at 10 seconds and 720p resolution by default — though under the Pro plan, that increases to 20 seconds and 1080p.
#4 - Sora Specific Prompting
According to OpenAI, Sora prompts should be around 120 words (2,500 characters max) and include these key elements:
Focus on one scene and one main action
Define a clear Subject + Action + Setting
Add simple camera or style cues (e.g. cinematic close-up)
(Optional) Include sound cues if needed
Here’s an example:
The bad news—unlike VEO-3, SORA doesn’t allow you to switch scenes mid-video.
However, you can still create multiple short clips and stitch them together later.
The good news—unlike VEO-3’s initial $2.50–$5 per video cost, Sora gives you 50 free videos every 24 hours.
In conclusion, Sora represents an intriguing step forward in AI video creation.
By framing it as a social app, OpenAI clearly aims to appeal to everyday users—unlike tools like Kling or Higgsfield that focus on cinematic-quality output for power users.
That said, there are growing rumors that OpenAI might take the classic VC route— attracting users with free access now, only to introduce paid tiers later (much like Uber or DoorDash did).
There are also rumors that OpenAI is struggling to handle the massive processing demands of all this free video generation—so their long-term strategy is still unclear.
For now, Sora videos are already going viral on Instagram (I don’t use TikTok), so our take is simple: enjoy it while it lasts!
Catch you next time
