County winter storm prep video: three 8-second clips for $3.70
A county or city office can turn Ready.gov's winter storm steps into three captioned vertical clips on Sume for about $3.70, with the alert terms spelled out.

What should a county winter storm prep video say?
Say what Ready.gov's winter weather page already says, in short pieces: seal the home against the cold, be ready to stay put without power, and keep a kit in the car. On Sume that is three 8-second vertical clips from Wan 3.0 at 720p, the official wording burned on as caption cues, and one Timeline render to join them. The maximum bill is about $3.70, and the first draft is a lunch-hour job.
This plan suits an emergency management office, a city public information officer or a township with no video staff. It does not replace an alert. The video explains what to do before a storm; the National Weather Service and your county's warning system still decide what is happening now.
Which Ready.gov lines make good clips?
Ready.gov lists steps to take before winter storms and defines three alert terms. Three groups of steps fit eight seconds each. Copy the wording into your cues yourself; do not ask the video model to render text, because generated lettering drifts.
| Clip | Ready.gov step | Caption cue to type |
|---|---|---|
| 1. The house | Insulation, caulking and weather stripping; learn how to prevent frozen pipes | Seal drafts. Know how to keep pipes from freezing. |
| 2. The power | Working smoke alarms and carbon monoxide detectors with battery backup; supplies for several days without power; extra batteries | Test your alarms. Stock supplies for several days. |
| 3. The car | Jumper cables, sand, flashlight, warm clothing, blankets, water, snacks and a full gas tank | Pack the car kit. Keep the tank full. |
How do you generate the three clips?
Each clip is one call to the asynchronous /v1/videos endpoint. Wan 3.0 accepts 2 to 30 seconds per the video generation docs. Setting generate_audio to false keeps the clips silent so the captions carry the message and nobody has to turn the sound on. Video models need a paid plan: Free is limited to image generation, and Pro is $40 a month with 4 concurrent jobs, so all three clips can run at once if you submit them in parallel.
import os, time, requests
H = {"Authorization": f"Bearer {os.environ['SUME_API_KEY']}"}
PROMPTS = [
"Frost on a window and a hand pressing weather stripping along a door frame",
"A glowing battery-backup smoke alarm and a flashlight on a dark kitchen counter",
"A car trunk with blankets, water, a flashlight and a bag of sand, snow falling outside",
]
def run(prompt):
r = requests.post("https://api.sume.com/v1/videos", headers=H, timeout=60, json={
"model": "wan-3.0", "prompt": prompt, "duration": 8,
"resolution": "720p", "aspect_ratio": "9:16", "generate_audio": False})
r.raise_for_status()
url = r.json()["polling_url"]
while True:
time.sleep(20)
s = requests.get(url, headers=H, timeout=60).json()
if s["status"] == "completed":
return s["unsigned_urls"][0]
if s["status"] == "failed":
raise RuntimeError(s.get("error", "generation failed"))
for p in PROMPTS:
print(run(p))How do you add the words and join the clips?
Send each finished clip's public HTTPS URL to POST /v1/video-captions with cues, each a text, start and end in seconds. Cues burn authored text without speech-to-text, which is the right path for silent footage. Add the three alert terms as a fourth text card if you want them: Ready.gov says a Winter Storm Watch comes 12 to 48 hours ahead, a Warning 12 to 24 hours before conditions arrive, and an Advisory covers snow and ice that can become life-threatening without precautions.
Then call POST /v1/timeline-1.0/plan. It is unbilled and returns the output duration and billable minutes before you spend anything. A 24-second render bills one minute at most. Render it with audio.mode set to silence and the three captioned clips in video[]; the Timeline docs list the fields.
What does it cost on Sume?
Wan 3.0 is priced per output second by resolution in the catalog. The rest comes from the API rate card; dollar figures use the roughly 66 credits to the dollar implied by the card's Timeline line (6.6 credits for $0.10).
| Item | Unit price | Quantity | Subtotal |
|---|---|---|---|
| Wan 3.0 clip, 720p, 8 s | $0.125 per second | 3 | $3.00 |
| Caption job with cues, up to 60 s | $0.20 per job | 3 | $0.60 |
| Timeline render, ceiling | 6.6 credits (about $0.10) per output minute | 1 minute | $0.10 |
| Total ceiling | $3.70 |
What should staff check before it goes out?
Watch every second for road, snow and vehicle details that look wrong for your region; regenerate any clip that shows the wrong kind of plow or a flooded street. Type phone numbers, shelter addresses and website names in the caption cues, never in the prompt. Keep the scope to preparation, since Ready.gov's page is general guidance, and add a link to your own warming-center list. If your budget is tight, 480p halves the clip line to $1.50 for the set; check the first draft on a phone before you commit to it.
Sources
Related posts
More in Use cases
- Cyber 5 2026 is five days: five 8-second clips for $5.00 on Sume
Cyber 5 runs Thanksgiving Thursday Nov 26 to Cyber Monday Nov 30. One 8 s 720p Gemini Omni Flash clip a day costs $1.00 each on Sume, $5.00 for all five.
- Cyber Monday Nov 30, 2026: $15.1B forecast, so queue recuts by Sunday
Adobe forecasts $15.1 billion in US online sales on Cyber Monday, Nov 30, 2026. Queue the recuts as one Sume bulk run by Sunday, Nov 29, and cap the spend.
- Slow down for phone numbers: phrase pacing in a Sume TTS script
Decagon says Chord's pacing follows each sentence. On Sume, speed is one setting per job, so render a phone number as its own slower take and join the takes.
- Dev Diwali Nov 24, 2026: a diya photo to an 8-second clip for $1.00
Dev Diwali is on Nov 24, 2026 on most calendars. One diya or ghat photo becomes an 8-second 9:16 Sume clip with Gemini Omni Flash 1.1 for $1.00 at 720p.
Written by Sume