ChabadLabs AI FOR SHLICHUS

Sora (OpenAI)

OpenAI's text-to-video model; the chat's late-2025 video-gen workhorse, since discontinued.

On this page

OpenAI's video-generation model. The chat's main video-gen workhorse in late 2025, partially superseded by Grok and Kling.ai by 2026.

External context: Sora has since been discontinued. This page is preserved because the chat's prompting lessons, especially describing clothing and setting to get community-accurate people, carry over to Grok, Kling, Veo, and other current video models.

What it is

sora.chatgpt.com (desktop) and a mobile app. It generates short video clips from prompts, natively under 8 seconds on mobile and longer on desktop. Invite codes were traded heavily in the chat through October and November 2025.

What the chat uses it for

  • Fundraising campaign videos, short narrative clips.
  • Holiday social-media content.
  • Educational videos, with shluchim demonstrating use for kids' programming.

Depicting your community accurately

The most useful prompting lesson from the Sora thread: to get images and video that reflect your community, describe attire, setting, and activity rather than identity or ethnicity labels. Concrete visual descriptions produce more accurate and more consistent results. For example:

"Women wearing modern modest dresses, hair visible, celebrating at a Shabbat dinner table."

"Women in contemporary casual clothing, smiling and talking while lighting Chanukah candles."

"A group of friends wearing smart casual party outfits at a festive Jewish event."

The fuller discussion lives in Prompting AI Image Tools to Depict Your Community, and the same approach applies to still-image generation. See Master Prompts for ready-to-paste examples.

Strengths

  • Quality high enough for social-media use.
  • The desktop version exceeds the mobile 8-second cap.
  • Integrated with your ChatGPT account.

Weaknesses

  • Length limits on mobile.
  • Invite-only for stretches.
  • Output quality varies by run; expect to generate several and pick the best.

Migration patterns

  • One user transitioned from Sora to Grok for educational videos.
  • Kling.ai for animated videos longer than 10 seconds.
  • Veo (Google) mentioned as an emerging alternative.
  • Synthesia, D-ID, and HeyGen for talking-avatar use.

Last updated May 2026 · Maintained by Hermes AI Agent