Video API · YouTube link or upload
Summarize any video, with timestamps, a transcript and an AI video prompt
Paste a public YouTube link or upload a clip. Picture and sound are analysed together, and you get a summary, a timestamped scene description, a transcript and a prompt to recreate the look in Veo, Sora, Runway or Kling.
The demo analyses the first 60 seconds. With an account, free tries cover the first 5 minutes and paid calls up to 20 minutes.
- Describes what is shown, not only what is said
- Timestamps in MM:SS, transcripts in the original language
- €0.05 per started minute, once per video, however many outputs you want
The problem
Too long to watch? Get what is in it.
You need the gist of a 20-minute video, the moment where something happens, a transcript you can search, or a prompt that recreates a clip's look. Tools that only read a video's captions miss everything that is shown and not said, and give up when there are no captions.
Captions miss the picture
Finding one moment means scrubbing
Recreating a look is guesswork
How it works
One pass over picture and sound
The video is analysed by a multimodal model that watches frames and listens to the audio at the same time, so the description, the transcript and the timestamps agree with each other.
Send a link or a file
A public YouTube URL is fetched by the model provider; a file goes through a direct upload.Choose the outputs
Any of summary, description, transcript and prompt. The price is the same for one or all four.Read, export or poll
Wait for the result in the same request, or poll the job. Export transcripts as SRT or VTT.
Real output
What you get, on a real video
The demo's output for the first 60 seconds of Goddard Earth Science Overview (NASA Goddard Space Flight Center, public domain), unedited. Notice that the prompt describes people without naming them.
Historic archival footage depicts the December 1968 launch of the Apollo 8 mission and the iconic Earthrise photograph taken by its crew. NASA official Christa Peters-Lidard explains how this perspective fostered global environmental awareness and inspired Earth Day. She introduces NASA Goddard Space Flight Center's Earth Science Division as a premier global hub for Earth scientists. Scenes then highlight Goddard scientists analyzing global climate and environmental data in collaborative meetings.
- 00:00–00:08Archival footage of the Saturn V rocket engines firing and lifting off from the launch pad amidst billowing smoke.
- 00:08–00:16Tracking shot of the Saturn V rocket ascending into the blue sky.
- 00:16–00:23Christa Peters-Lidard speaks in front of a digital composite showing the lunar surface and Earth.
- 00:23–00:28Archival black-and-white footage of Apollo 8 astronauts operating equipment inside the spacecraft cabin.
- 00:28–00:38Still photograph of the iconic Earthrise view showing Earth above the lunar horizon in space.
- 00:38–00:42Outdoor entrance of NASA Goddard Space Flight Center showing flagpoles and the monument sign.
- 00:42–00:48Christa Peters-Lidard speaking to the camera in front of the lunar background.
- 00:48–00:52A conference room where a researcher presents global satellite data on a monitor to seated colleagues.
- 00:52–00:55Close-up profile of Christa Peters-Lidard listening attentively in a meeting.
- 00:55–01:01A young male scientist gestures toward a monitor displaying global wildfire analysis charts.
- 00:00–00:07Launch Control: Three, two, one, zero. We have commit. We have liftoff. The clock has run. We have liftoff.
- 00:08–00:23Christa Peters-Lidard: Over 50 years ago, the Apollo 8 mission launched to the Moon in December of 1968. As part of this mission, three astronauts were able to capture an iconic image of the Earth rising over the Moon, known as Earthrise.
- 00:24–00:28Astronaut: You got that launch... going around to a good view of the Earth.
- 00:28–00:37Christa Peters-Lidard: This image inspired a generation of environmental activism, starting with Earth Day and decades of environmental activism since.
- 00:40–01:00Christa Peters-Lidard: The NASA Goddard Space Flight Center's Earth Science Division is the largest collection of Earth scientists on the planet. And our job is to be the nation's trusted source of comprehensive environmental information about the current state and the future of the Earth.
Documentary montage featuring vintage 1960s Apollo rocket launch footage, the iconic Earthrise view over the Moon, an interview with a female scientist in a blue blazer against a space backdrop, exterior shots of NASA Goddard facility, and contemporary scientists collaborating over global environmental maps in a modern office. Authentic documentary style with archival film grain transitioning to crisp digital footage, dramatic and inspiring tone, accompanied by historic radio chatter and clear narrative voiceover. Style: documentary archival and digital video Aspect ratio: 16:9 Negative prompt: blurry, low quality, CGI artifacts, distorted faces, unrealistic motion, modern elements in archival scenes
- 00:00–00:08Vintage 16mm film capture of a Saturn V rocket ignition, heavy flame and smoke billowing from pad thrusters as launch tower arms retract.
- 00:08–00:16Archival telephoto shot of Saturn V rocket climbing into the sky propelled by bright fiery exhaust plume.
- 00:16–00:23Medium shot of a professional woman with blonde curly hair wearing a blue blazer speaking to camera in a studio with Earthrise graphic background.
- 00:23–00:28Black and white grainy footage inside an Apollo capsule of astronauts floating and maneuvering instruments.
- 00:28–00:38Cinematic high-detail recreation of Earthrise showing the blue marble Earth rising above the barren, cratered gray lunar surface in deep black space.
- 00:38–00:42Medium exterior pan from fluttering American and state flags down to the stone entrance sign of NASA Goddard Space Flight Center under a partly cloudy sky.
- 00:42–00:48Medium shot of female scientist in blue blazer continuing interview against lunar horizon background.
- 00:48–00:52View looking through an open office doorway at scientists sitting around a conference table while one stands beside a screen displaying Earth data.
- 00:52–00:55Close-up side profile of a blonde businesswoman listening intently during a team presentation in an office.
- 00:55–01:01Medium shot of a young researcher pointing and presenting global wildfire satellite heat maps displayed on a large wall monitor.
Who it is for
Teams that work with a lot of video
Content and social teams
Course and edtech platforms
Research and media monitoring
AI video creators
Developers
Accessibility
For developers
One request, four outputs
Full reference in the API docs. Guides: transcripts and summaries from YouTube and video to prompt.
# Summary, timestamped scenes, transcript and an AI video prompt, in one job
curl -X POST https://api.opencorpo.com/v1/video/analyze \
-H "X-API-Key: $OPENCORPO_API_KEY" \
-H "Content-Type: application/json" \
-d '{
"source": {"type": "youtube", "url": "https://www.youtube.com/watch?v=opjOMXqjWdI"},
"outputs": ["summary", "description", "transcript", "prompt"],
"wait": true
}'
# Subtitles from the transcript
curl -o talk.srt "https://api.opencorpo.com/v1/video/jobs/vj_.../export?format=srt" \
-H "X-API-Key: $OPENCORPO_API_KEY"MCP server
Use it from Claude or ChatGPT
A YouTube MCP server for Claude and ChatGPT: paste a link in the chat and get the summary, the key moments with timestamps and the transcript, even when the video has no captions. The same account, free tries and balance as the API, in Claude, ChatGPT, Claude Code, Cursor and VS Code.
- One URL:
https://api.opencorpo.com/mcp - Tools:
analyze_video,get_video_job - Sign in once with your OpenCorpo account. Disconnect any app at any time.
Ask, for example
Pricing
Pay per minute of video. No subscription.
| Operation | Price |
|---|---|
| Free tries with an account | 3 free |
| Video analysis | €0.05 / minute |
What that comes to
- A 3-minute clip€0.15
- A 10-minute video€0.50
- A 20-minute talk€1.00
- 100 ten-minute videos€50.00
- Billed per started minute of analysed video, once per video, with all four outputs included. The quality model costs ×3, high detail ×2.
- 3 free analyses per account, each covering the first 5 minutes of a video.
- The cost of the longest analysis your balance allows is reserved when a job starts, and only the actual length is charged. Failed jobs cost nothing.
Every API on the pricing page. Prices exclude VAT where applicable.
FAQ
Questions about the video summarizer
How do I summarize a YouTube video?
Paste a public YouTube link into the demo on this page (the first 60 seconds, no account), or send it to POST /v1/video/analyze with your API key. You get a one-paragraph summary, a scene-by-scene description with MM:SS timestamps, the transcript and an AI video prompt from one analysis.
Does it work with my own video files?
Yes. Upload MP4, MOV, WebM and other common formats, up to 500 MB through the uploads endpoint, and up to 20 minutes are analysed per video. The demo here takes short clips.
Do you download YouTube videos?
No. The YouTube link is handed to the model provider, Google, which fetches the video on its side. That is why it only works with public videos, and why nothing is scraped or stored by OpenCorpo.
What languages are supported?
Speech is transcribed in the language spoken and never translated; you can pass a language hint such as fr. Each line of the transcript is timestamped and attributed to a speaker where the video makes that possible.
What is the AI video prompt?
A text-to-video brief for recreating a similar video in Veo, Sora, Runway, Kling and the like: one prompt for the whole video, a style line, the aspect ratio, a negative prompt, and one self-contained prompt per shot, because generators remember nothing between clips. It describes what is seen and never names real people.
Can I get subtitles?
Yes. Export the transcript (or the scene description) as SRT, VTT, plain text or JSON with GET /v1/video/jobs/{id}/export.
How accurate are the timestamps?
Timestamps come from the model's own reading of the video, which by default looks at about one frame per second together with the full audio track. Expect them to be right to within a second or two; for fast cuts you can raise the frame rate.
Who processes the video?
Google's Gemini API, on a paid plan, so Google does not use the content to improve its models. Processing is not guaranteed to happen in the EU. Uploaded files are deleted from Google after the analysis, and demo uploads here are deleted within 24 hours.
How much does it cost?
€0.05 per started minute of analysed video, charged once per video whatever outputs you ask for. Every account gets 3 free analyses (the first 5 minutes of each), then pays from a prepaid balance with no subscription.
Guides
Summarize your videos from your own code
Create an account, get an API key and use 3 free analyses. No card needed until you top up.