Same models. Most for less.
And we make the video.

A presenter-led product ad — shot list, native lip-synced dialogue, your product in every shot, music under the cut. No shoot, no editor: two calls to a finished ad.

1. 1

   Send the brief.

   API docs

   One POST — a 24-second vertical product ad: three shots, your real product in frame, music under the cut, captions burned in. The AI writes it, renders it, cuts it. No questionnaire, no shoot.

   Request

   curl -X POST https://api.turboclip.ai/v1/ad \
     -H "Content-Type: application/json" \
     -H "x-api-key: tc_sk_..." \
     -d '{
       "brief": "Launch ad for the Ridgeline Pro hiking jacket. Follow one hiker from a dark trailhead to a sunrise summit. Waterproof, featherlight, built for cold alpine starts. End on the jacket and the tagline: Made for the way up.",
       "settings": {
         "adType": "product",
         "videoModel": "gemini-omni-1.1-flash",
         "aspectRatio": "9:16",
         "duration": 24,
         "shots": 3
       },
       "referenceImageUrls": ["https://cdn.ridgelinegear.com/ridgeline-pro/product.jpg"],
       "music": true,
       "captions": true
     }'

   Copy

   Response · 200

   { "success": true, "data": { "projectId": "3f2a1b4c-…", "jobId": "9b1c7e2d-…",
       "processType": "full_pipeline_video", "creditsReserved": 105, "creditsRemaining": 895 } }

   Output AI Ad · 9:16

   

2. 2

   Poll one URL.

   That’s the whole loop — no render call, no second endpoint. When status flips to completed, the finished MP4 is right there in the same response, music already under the cut.

   Request

   curl https://api.turboclip.ai/v1/jobs/{jobId} \
     -H "x-api-key: tc_sk_..."

   Copy

   Response · 200

   { "success": true, "data": { "status": "completed", "creditsConsumed": 105,
       "output": { "type": "video", "videoUrl": "https://…/ridgeline-pro-launch.mp4" } } }