2026-11-16 · 5 min read

How to add captions to interview videos with an API

A film strip of interview scenes, the middle frame showing the captioned word TALK in neon
This image was generated with AI.

Interview content is the voice: founder talks, expert conversations, customer stories, podcast video. The words are the entire product, and a large share of the audience watches with the sound off or half-listening while they work. caption.sh adds the captions with an API call instead of an editing session.

The workflow

import requests, time

API = "https://api.caption.sh/v1"
HEADERS = {"Authorization": "Bearer YOUR_API_KEY"}

job = requests.post(
    f"{API}/videos",
    headers=HEADERS,
    json={
        "source_url": "https://your-cdn.example/founder-interview.mp4",
        "options": {
            "font": "inter",
            "position": "bottom",
            "max_words": 5,
            "max_lines": 2,
            "caps": False,
            "backdrop_opacity": 0.35,
        },
    },
).json()
job_id = job["job_id"]

for _ in range(120):
    job = requests.get(f"{API}/jobs/{job_id}", headers=HEADERS).json()
    if job["status"] in ("done", "error"):
        break
    time.sleep(5)

if job["status"] == "done":
    r = requests.get(f"{API}/jobs/{job_id}/result", headers=HEADERS,
                     allow_redirects=False)
    video_url = r.headers["Location"]

Jobs run queued, awaiting_upload, running, then done or error; poll until a terminal status. The result endpoint answers 302, so read the Location header with redirects disabled. Results are stored for 24 hours, then deleted.

Styling for talking heads

  • Interview pacing is conversational: max_words at 5 with caps off reads like dialogue instead of a hype reel.
  • Bottom position with a modest backdrop_opacity keeps text readable over changing backgrounds without covering the speaker.
  • For two-camera interviews, keep captions consistent across cuts so the eye does not re-find the text every switch.

Accuracy review

Names of people, companies, and products are where any transcription earns its review pass, and interviews are dense with all three. The generated captions give you a timed draft to check against rather than a transcript to type from scratch. For social cuts, verify the guest's name and title before posting.

Tradeoffs and cost

  • Burned-in captions cannot be toggled, resized, or translated. Keep platform captions enabled for viewers who need adjustable text.
  • Each job covers up to 10 minutes. Full-length interviews should be segmented, which also produces your social cuts.
  • A 429 means the transcription throughput ceiling is reached. Wait and retry rather than submitting duplicates.
  • At $0.10 per minute prepaid, captioning a 10-minute segment costs $1.00.

Back to all articles