Canva MP4 export quality is now target pixels: caption it on Sume

The answer

Since September 18, Canva's MP4 export quality is a target pixel count and the design keeps its own aspect ratio, with no letterboxing. A portrait design exported as horizontal_4k therefore stays portrait at 2160x3840, and the horizontal and vertical variants return identical pixels. You choose a size, not an orientation.

Canva exports silent, uncaptioned motion unless you built the captions into the design. For a talking clip, a Sume video-captions job can add them afterwards from the exported file.

Canva export facts

From the create design export job reference and the changelog.

Add captions on Sume

POST /v1/video-captions takes a public HTTPS video_url and a style: slam, punch, tiktok-green or korean-ad. Host the Canva export where Sume can fetch it over HTTPS without a signed or private link, then submit.

Captions read the speech in the file. A silent clip fails with caption_no_speech, so for authored text pass cues yourself instead of relying on transcription.

curl -X POST https://api.sume.com/v1/video-captions \
  -H "Authorization: Bearer $SUME_API_KEY" \
  -H "Idempotency-Key: caption-2026-10-04-001" \
  -H "Content-Type: application/json" \
  -d '{"video_url":"https://example.com/exports/promo.mp4","style":"tiktok-green"}'

Caveats

The Sume wire details are in the captions docs alongside the job model.

• Pick the export size from the destination: 1080p is plenty for a feed, and 4k portrait is large to upload.
• Poll the job, do not assume it is done: statuses are queued, processing, completed, failed and canceled.
• Check the Canva export URL lifetime in its own docs before you queue the Sume call.