Launchscene
PricingSign in
Explore

Meet Jockey, TwelveLabs' video intelligence agent, now in general availability

TwelveLabs · Oct 2026 · 0:41 · AI-reviewed

00:00:00 / 00:41:22

Jockey announces its arrival in ChatGPT and Claude with short kinetic title cards, then stages a fast ChatGPT conversation: analyze trending forest clips, retrieve matching footage, build a reel, swap a clip and generate three variants. Typed prompts, an orange cursor, floating video cards and reframed interface views drive the demonstration. Jockey and Powered by TwelveLabs close the film over a lively music bed with a brief countdown sample.

No filming indicatedEstimated needs: 1 capture group — your screens or recordings · 1 official logo — reused across the film

More like this

Meet Jockey, TwelveLabs' video intelligence agent, now in general availability, scene by scene

· 21 scenes as text

  1. 01 · Product intro · 0:00–0:01

    Centered kinetic words build Jockey has entered the chat on a pale peach-to-white field. Jockey begins coral, the connecting words are black, and chat is coral.

    On-screen text: “Jockey has entered”, “the”, “chat”

    Techniques: Kinetic type

    Camera: Static

  2. 02 · Product intro · 0:01–0:02

    The black ChatGPT symbol and wordmark hold centered alone on the white title card.

    On-screen text: “ChatGPT”

    Camera: Static

  3. 03 · Product intro · 0:02–0:03

    A separate centered Claude lockup, with coral starburst and black serif wordmark, replaces ChatGPT on white.

    On-screen text: “Claude”

    Camera: Static

  4. 04 · Demo · 0:03–0:05

    The ChatGPT composer replaces the platform card. A Jockey chip appears, the prompt Hey Jockey, I have 10 minutes to create a viral video types in, and an orange cursor clicks send, turning it into a peach user bubble.

    On-screen text: “Hey Jockey, I have 10 minutes to create a viral video”, “Jockey”

    Techniques: Kinetic type, Cursor demo

    Camera: Static

  5. 05 · Demo · 0:05–0:07

    A rapid push into the chat response reveals the large line Let’s get started! beneath the cropped peach user message. The response holds in the enlarged framing.

    On-screen text: “Let’s get started!”

    Techniques: UI zoom

    Camera: Push in

  6. 06 · Demo · 0:07–0:10

    A hard edit returns to the full ChatGPT composer. The prompt Here are some trending videos. Why are they doing well? types with three Instagram links while forest thumbnails slide around the input and the orange cursor clicks send.

    On-screen text: “Here are some trending videos. Why are they doing well?”, “Jockey”

    Techniques: Kinetic type, Cursor demo

    Camera: Static

  7. 07 · Demo · 0:10–0:13

    Forest video cards float into an overlapping cluster with the word Analyzing. A bright dissolve reveals the results, and the same answer scrolls continuously through Trending location, Strong opening hook and Satisfying payoffs; these are three analysis reasons, not separate shots at12379ms.

    On-screen text: “Analyzing”, “Trending location”, “Strong opening hook”, “Satisfying payoffs”

    Techniques: Kinetic type, UI zoom

    Camera: Static

  8. 08 · Demo · 0:13–0:16

    The scrolled analysis dissolves into a composer surrounded by three forest clips. The user types Please tell me we also have some footage like this somewhere in our catalog of 1,347 videos?! and the orange cursor sends it.

    On-screen text: “Please tell me we also have some footage like this somewhere in our catalog of 1,347 videos?!”, “Jockey”

    Techniques: Kinetic type, Cursor demo

    Camera: Static

  9. 09 · Demo · 0:16–0:18

    Through a bright dissolve, a rounded gallery assembles forest clips: a walker at left and redwood views beside it. The reply types You do. Right here. And eight more like it. The view starts pushing into the same gallery.

    On-screen text: “You do. Right here. And eight more like it.”

    Techniques: UI zoom, Kinetic type

    Camera: Push in

  10. 10 · Demo · 0:18–0:21

    A continuous push and upward reframe of the same forest gallery makes room for a composer. Okay what should we make? types into the input beside the Jockey chip; the gallery remains above.

    On-screen text: “Okay what should we make?”, “Jockey”

    Techniques: UI zoom, Kinetic type

    Camera: Push in

  11. 11 · Demo · 0:21–0:23

    The peach user bubble reads Okay what should we make? while the assistant types A 15-second reel. Lead with the reveal, then the best reactions. The Jockey composer begins entering below.

    On-screen text: “Okay what should we make?”, “A 15-second reel. Lead with the reveal, then the best reactions.”

    Techniques: Kinetic type

    Camera: Static

  12. 12 · Demo · 0:23–0:25

    A close framing tracks across the input as Sounds like a plan, let’s make it! types in. The orange cursor clicks the black send arrow beside the Jockey chip.

    On-screen text: “Sounds like a plan, let’s make it!”, “Jockey”

    Techniques: UI zoom, Cursor demo, Kinetic type

    Camera: Pan

  13. 13 · Demo · 0:25–0:26

    Floating forest-video cards assemble into a tight vertical stack and resolve into the large vertical preview with a five-clip filmstrip. This is an interface animation; the clipped text from the preceding composer is not a new title.

    Techniques: UI zoom

    Camera: Push in

  14. 14 · Demo · 0:26–0:29

    The same vertical clip shows a cap-wearing person in plaid walking through redwoods. The five-clip filmstrip fades and the view reframes downward into a composer where Swap the last clip with the one before types. The walking shot remains continuous across28261ms.

    On-screen text: “Swap the last clip with the one before”, “Jockey”

    Techniques: Kinetic type, UI zoom

    Camera: Static

  15. 15 · Demo · 0:29–0:31

    The vertical preview hard-cuts to a carved SEQUOIA NATIONAL PARK sign. The swap request is sent, its composer fades, and the five-clip timeline returns below the same sign.

    On-screen text: “SEQUOIA NATIONAL PARK”, “Swap the last clip with the one before”

    Techniques: UI zoom

    Camera: Pull out

  16. 16 · Demo · 0:31–0:32

    The large vertical preview switches from the park sign to sunlit redwood trunks while the five-clip timeline stays below.

    Camera: Static

  17. 17 · Demo · 0:32–0:33

    The large vertical preview switches to a person standing in a bright grassy clearing beside dark forest. The timeline remains and the preview scales down before the next input.

    Techniques: UI zoom

    Camera: Pull out

  18. 18 · Demo · 0:33–0:36

    A close composer types Create me three variants of this video that I can send to the team for approval. The orange cursor enters and clicks the send button.

    On-screen text: “Create me three variants of this video that I can send to the team for approval”, “Jockey”

    Techniques: Kinetic type, Cursor demo

    Camera: Static

  19. 19 · Demo · 0:36–0:38

    A bright dissolve reveals three vertical forest-video variant cards with play icons and MP4 filenames. The assistant reply types Here are three variants of the reel, ready for review.

    On-screen text: “Here are three variants of the reel, ready for review.”

    Techniques: Kinetic type

    Camera: Static

  20. 20 · Outro · 0:38–0:39

    The coral-gradient Jockey wordmark appears softly on white, sharpens and holds centered. No call to action is written.

    On-screen text: “Jockey”

    Techniques: Kinetic type

    Camera: Static

  21. 21 · Outro · 0:39–0:41

    A hard cut replaces Jockey with the centered black TwelveLabs pixel-horse symbol and wordmark beneath Powered by. The white end card holds.

    On-screen text: “Powered by”, “TwelveLabs”

    Camera: Static