How teachers can create classroom videos with Google Flow and Vids

Google Flow and Google Vids overlap, but they solve different parts of the job. Flow is useful when you need to generate and refine a specific visual movement using text, first-and-last frames, ingredients, or another video. Vids is the scene-based editor where you can combine those clips with slides, images, narration, captions, recordings, and questions.

For a teacher, the goal is not producing a long AI film. It is making one process, movement, transition, or imagined situation easier to see and discuss, then placing that moment inside a clear instructional sequence.

Plan for the teacher to create, review, and download the material. Google currently requires Flow users to be at least 18 and availability differs by country or region. School administrators can also control access to Vids and its Gemini features. Learners can watch the finished MP4 on the classroom smart board without needing their own Flow or Vids accounts.

Eligible users without a Google AI subscription may receive free Flow credits, while Google AI Plus, Pro, and Ultra plans provide different allowances and features. Credit costs depend on the selected model and generation settings, so check the value shown inside Flow before generating. Free, Plus, and Pro Veo videos may include a visible AI watermark.

The basic Vids editor is available with a Google Account in supported locations, while AI video generation, avatars, voiceovers, and other Gemini features depend on the account or Workspace plan. Teachers using a school account should first check which features their administrator has enabled.

Begin with a moment that is difficult, unsafe, too slow, too fast, too small, or impossible to demonstrate in the room. Reduce it to one short visual event. A six-second clip of a ball changing direction is more teachable than a two-minute AI film about the laws of motion.

Write the question learners will answer after watching before you generate anything. This prevents the video from becoming decoration.

Open Flow on a computer and create one project for the lesson. In the generation settings, choose Video and confirm the active model, aspect ratio, clip length, number of outputs, and credit cost. For most classroom displays, landscape 16:9 is the practical starting point.

Generate one output first. More variations use more credits and make review slower. Keep each concept in its own project or clearly named collection so you can find the approved clip later.

A classroom prompt should describe the visible evidence, not only the topic. “Create a video about force” leaves nearly every decision to the model. Describe the subject, action, environment, camera, visual treatment, timing, and anything that must not appear.

Keep the camera instruction simple. A fixed side view often communicates classroom motion more clearly than a cinematic orbit, zoom, or rapid cut.

Suppose you want learners to observe how contact changes a football's motion. Prepare two rights-cleared images with a similar player, kit, camera angle, lighting, and background. The first should show the player approaching a stationary ball. The second should show the finishing pose with the ball moving away.

You may use Google Images to discover possible sources, but an image appearing in search results does not give permission to reuse it. Prefer your own photographs, school-approved assets, public-domain material, Creative Commons images whose licence permits your use, or properly licensed stock. Avoid identifiable children unless your school has explicit permission.

If your player and background are separate assets, first create two clean composite images: a starting frame and an ending frame. This gives Flow a clear visual destination. In Flow, choose Video, then Frames, and add them as the start and end frames.

Frames control where a clip begins and ends. Ingredients are visual references that help Flow reuse a character, object, location, or style. Use them when you need the same football, laboratory setup, historical setting, or illustrated character across several clips.

Clean references work better than busy pictures. Place the main subject on a plain or clearly separated background where possible, keep the visual style consistent, and explicitly tell Flow what each reference represents.

Use Flow to create the motion-rich visual, then use Vids to build the teachable sequence around it. Download the approved Flow clip and open vids.new. Start with a landscape Vid for a typical smart board, then upload the Flow clip into its own scene.

A simple classroom Vid can use three scenes: a prediction question before the clip, the generated clip itself, and an explanation or transfer question afterward. This structure keeps the video connected to learning rather than turning it into passive viewing.

You can also convert an existing Google Slides presentation into Vids. Each slide becomes a scene and its speaker notes can become the starting script. Replace only the scene that needs motion with the approved Flow clip.

If your account includes AI video clips, Vids can generate a clip without leaving the editor. Open AI video from the right sidebar, choose Create, describe the scene, and optionally upload Ingredients. Google currently allows up to seven reference images for a generated Vids clip, though limits and availability can change.

Use this route when the generated clip is one part of a larger Vid and you want to stay inside the same timeline. Use Flow when you need more deliberate frame control, repeated visual consistency, scene extension, or deeper clip refinement.

The football setup shown below uses two ball references and a common field background. The prompt must tell Vids what each image represents and what relationship or motion it should demonstrate.

Once the visual sequence is correct, add only the support learners need. Use concise on-screen questions, a teacher recording, an available AI voiceover, or captions. Avoid narrating every visible detail; direct attention to the evidence that matters.

Vids behaves more like a collaborative presentation editor than a complex film editor. You can change scene timing, animate text or images, add stock or uploaded media, record your screen or camera, invite a colleague to comment, and reuse the file for another class by making a copy.

Use Flow where motion or change carries the meaning. The generated clip should lead into explanation, comparison, or practice—not replace them.

Watch every generated frame before showing the class. A polished clip can still contain a wrong sequence, implausible movement, misleading scale, invented text, stereotypes, or distracting visual errors.

Regenerate by changing one variable at a time. If the motion is correct but the camera moves, ask Flow to lock the camera and preserve everything else. This is faster and more reliable than rewriting the entire prompt.

If you are staying in Flow, download one approved clip from its asset menu or use Scenebuilder to arrange and trim several clips before downloading the scene. If you assembled the lesson in Google Vids, use File and Download as MP4, or export the finished video to Google Drive.

Test the downloaded file on the classroom computer before the lesson. Keep a local copy so the activity does not depend on internet access. Put the observation question on the slide before the video, play the clip once without discussion, then replay and pause at the decision point.