{
  "$type": "site.standard.document",
  "bskyPostRef": {
    "cid": "bafyreiaj6y6vtnzeewiif3kkapwwflvlypsa42rlx5rl33ihysdh4zjymm",
    "uri": "at://did:plc:pgryn3ephfd2xgft23qokfzt/app.bsky.feed.post/3mkav2jxdmss2"
  },
  "path": "/t/how-to-create-an-ai-video-more-than-4-or-10-seconds-please/175531#post_2",
  "publishedAt": "2026-04-24T14:46:33.000Z",
  "site": "https://discuss.huggingface.co",
  "tags": [
    "https://help.runwayml.com/hc/en-us/articles/46974685288467-Creating-with-Gen-4-5",
    "https://help.runwayml.com/hc/en-us/articles/48324313115155",
    "https://www.adobe.com/products/firefly/features/image-to-video.html",
    "https://www.luma-ai.com/luma-dream-machine-v1-5/"
  ],
  "textContent": "* * *\n\nYes, this is possible. Most AI video tools generate short clips first (usually 5–10 seconds), then you **extend** or **continue** them until you reach 30–60 seconds.\n\n## Simple method (works in most tools)\n\n  1. Start with your image and generate the first short clip (5–10s).\n  2. Use **Extend / Continue** (if available), or export the **last frame** and use it as the next image input.\n  3. Keep the same prompt style so character, camera, and lighting stay consistent.\n  4. Repeat until total video length is around 30s or 1 minute.\n  5. Combine all clips in an editor (CapCut, DaVinci Resolve, Premiere), then add sound/music.\n\n\n\n## Good tools to try\n\n  1. **Runway** (very good for image-to-video + continuation)\n\n     * Gen-4.5 overview: https://help.runwayml.com/hc/en-us/articles/46974685288467-Creating-with-Gen-4-5\n     * Image-to-video prompting + continuation workflow: https://help.runwayml.com/hc/en-us/articles/48324313115155\n  2. **Adobe Firefly** (beginner-friendly interface)\n\n     * Image-to-video feature: https://www.adobe.com/products/firefly/features/image-to-video.html\n  3. **Luma Dream Machine**\n\n     * Extend workflow reference: https://www.luma-ai.com/luma-dream-machine-v1-5/\n\n\n\n## Quick quality tips\n\n  * Write one clear scene prompt and reuse it for each segment.\n  * Keep camera motion small for better continuity.\n  * If faces drift, reduce motion and use the same reference image style each step.\n  * Generate extra takes and pick the best segments before stitching.\n\n\n\nfor example here are some ready-to-use sample prompt template for cinematic, realistic, or anime style videos.\n\nUse this copy/paste template:\n\n`[Subject] in [environment], [time of day], [lighting], cinematic [camera shot], [camera movement], [action], realistic details, natural motion, high consistency, keep same character design, same outfit, same face, same background style, smooth transition from previous clip.`\n\nExample (realistic):\n`A young woman in a red coat standing on a rainy Tokyo street at night, neon reflections on wet pavement, cinematic medium shot, slow dolly-in camera movement, she turns and smiles softly while traffic passes behind her, realistic skin texture, natural motion, high consistency, keep same character design, same outfit, same face, same background style, smooth transition from previous clip.`\n\nExample (anime):\n`Anime boy with silver hair on a windy cliff at sunset, golden rim light, dramatic wide shot, gentle orbit camera movement, cape flowing as he looks toward distant mountains, clean anime linework, vibrant colors, natural motion, high consistency, keep same character design, same outfit, same face, same background style, smooth transition from previous clip.`\n\nNegative prompt (if supported):\n`low quality, blurry face, extra fingers, distorted hands, warped body, flicker, frame jitter, inconsistent face, outfit change, background change, text, watermark, logo`\n\nFor longer videos:\n\n  1. First clip: use full prompt.\n  2. Next clips: reuse same prompt + add `continue from previous shot`.\n  3. Change only one thing at a time (camera OR action) for smoother continuity.\n\n\n\nGood luck brother.",
  "title": "How to create an AI Video more than 4 or 10 seconds please?"
}