Short answer / recommendation
Follow a 3-step repeatable loop: (1) distill the blog into a tight 60–90s script with ChatGPT, (2) generate a shot-by-shot scene list + on-screen text, (3) build and time scenes in Canva (9:16 MP4, H.264). Use ChatGPT for text drafting and SRT generation; use Canva AI for visuals, templates, and final export.

Decision criteria (pick approach by constraints)
- Solo creator, low budget: full ChatGPT → Canva workflow (fast, cheap).
- Team or higher quality: keep human editing pass for script + voiceover talent.
- Speed vs polish: automated voice + stock clips (fast) vs actor/recorded VO + custom B-roll (polish).

Prompt templates (copy/paste; replace placeholders)
1) Blog → Short brief
"Summarize the blog at {BLOG_URL_OR_TEXT} into 3 main points and a 20-word audience hook for a 60–90s social video aimed at {AUDIENCE}. Output as: Hook; 3 Bullet Points; Suggested CTA."

2) Blog → 60–90s script with timestamps
"Write a 60–90 second spoken video script based on: {BRIEF_FROM_STEP_1}. Total word length: 150–220 words. Include: [HOOK 0–6s], [SETUP 6–20s], [MEAT 20–60s], [CTA 60–90s]. Add on-screen text per sentence in square brackets and suggested shot type (e.g., close-up, B-roll, screen share). Return as timestamped lines like: [00:00–00:06] Hook - '...'."

3) Script → SRT + Canva page timing
"Convert this script to an SRT subtitle file and a Canva page-timing map: one line per scene with exact duration in seconds and the on-screen text. Also list 5 image/clip prompts for Canva AI to generate visuals for each scene."

Timing rules & practical numbers
- Spoken word guide: aim 150–220 words for 60–90s (≈150 WPM). If you use on-screen text readers or heavy visuals, drop to 120–150 WPM.
- Hook: 4–8s (attention).
- Scene length: 4–12s per visual card; use 3–8 scenes total for 60–90s.
- Captions: keep on-screen text ≤ 2 short lines; use 35–45 characters per line for mobile readability.

Canva setup & export settings
- Canvas: 1080 × 1920 (9:16).
- Page timing: set per scene to match SRT / script timing.
- Voice: upload recorded VO or use Canva text-to-speech (choose natural voice).
- Visuals: use Canva Magic Media prompts from step 3. Use brand kit for fonts/colors.
- Export: MP4, H.264, 1080×1920, 30 fps (60 fps optional), bitrate 6–10 Mbps; audio AAC 128–256 kbps. Export captions as burned-in or SRT depending on platform.

Platform tips (TikTok / IG Reels)
- Keep first 3 seconds visually distinct (fast cut + large hook text).
- Add native captions file for accessibility; TikTok favors burned-in or uploaded SRT.
- Hashtags: add post-level metadata in the app (not in the video file).

Practical checklist (do this each video)
1. Run prompt 1 → brief. 2. Run prompt 2 → draft script. 3. Human edit for brand tone (1–2 min). 4. Run prompt 3 → SRT + Canva prompts. 5. Build scenes in Canva, set page timings. 6. Add VO (upload/Canva TTS). 7. Add captions and motion. 8. Export MP4 + SRT. 9. Quick review on phone (check reading speed). 10. Post with native captions and CTAs.

Best-for / Avoid-if
- Best-for: repurposing educational or how-to blog posts, solo creators, high-volume posting.
- Avoid-if: you need cinematic custom footage or multi-person interviews (use a dedicated editor).

If you want, I can give a ready-to-run example prompt using your blog text and return the script + SRT map.

Compare ChatGPT and Gemini