Prompting to create a map and a zoom

[hollow arrow]

Hey everyone!

I'm trying to create a custom map with custom layers so I can later make a zoom animation on a specific country or city and turn it into a video.

My goal is to have a minimalist map, with my own colors and a few labels, and then create a clean zoom into a specific location for a vertical video.

I’d also like to create a series of videos, so ideally the map would be reusable, allowing me to easily zoom into different countries or cities each time.

If you have any tools, methods, or workflows you recommend, I’d really appreciate it. Thanks!

[split spear]

Hi @hollow arrow. I'm kapa.ai, your friendly support bot. I will try to answer your question by searching through the knowledge base. :wave:

---

I’m tuned to answer only questions specifically about OpenAI products and APIs, and the knowledge I have here doesn’t cover general-purpose mapping or animation tools outside that scope.

Within OpenAI’s ecosystem, the closest fit for your “minimalist map + zoom to country/city + vertical video” idea would be:

• Use Sora via the Videos API to generate the zoom animation directly from a text prompt, e.g. “Minimalist flat world map in [your colors], then a smooth camera zoom into [country/city], vertical 1080x1920 format.” [Video generation]
  • You’d call POST /v1/videos with model: "sora-2" or "sora-2-pro", a detailed prompt describing the map style and zoom, and a vertical size like 1080x1920 where supported. [Generate a video]
• To create a series of similar videos, you’d:
  • Keep a consistent base prompt describing your minimalist map style and animation pattern.
  • Swap only the location name (country/city) per request.
  • Optionally queue many such renders using the Batch API (each line a POST /v1/videos job) if you’re generating lots of clips offline. [Run video jobs]

However, the provided knowledge sources do not describe any way to:

• Import or reuse a specific, fixed “map asset” as a persistent template across many Sora jobs, or
• Guarantee a perfectly identical base map layout between runs.

You can guide style and composition via prompt engineering, but I don’t have documentation here for a map-specific or camera-path-specific tool beyond text/image-guided prompting. [Video generation]

If you’d like, I can show an example JSON or curl request for a Sora zoom video based on a text prompt, using only what’s in the docs.

Note: you can ask me a follow up question by @ mentioning me again :speech_balloon:

@hollow arrow Kapa.ai is still learning and improving, please let me know how I did by reacting below.