AI Video Generator:
Turn Text or a Photo Into a Clip

The AI video generator in Secret Chat AI makes a clip from a written scene or a picture. Choose a model, set the length and format, then describe what should happen. You can start from a blank frame, animate a photo, or make a picture from the AI Image Generator move.

There are two entries in the model list. Video has Kling, Veo, and Grok. The uncensored Video generator, listed as Video Lax with a flame, has Wan, Seedance, and MiniMax. Every video model is paid. The exact price appears before you start, and a failed clip costs nothing.

Text to Video or Image to Video AI

For text to video, leave both frame slots empty and write a clear scene. Name the subject, action, camera movement, setting, and the look you want. The aspect ratio controls the shape of the result. The default is vertical 9:16, though the available shapes depend on the model.

For image to video AI, place a picture in the First frame slot and describe the movement. The clip begins from that picture. Most models follow the picture's shape, while Veo keeps the aspect ratio chosen in the settings. The picture can be a photo, an illustration, or a result you downloaded from the Image Generator.

Some models also accept a Last frame. That slot opens after you add the first picture. It lets you give the clip a destination: a product can finish in a chosen position, a camera move can land on a planned view, or one still image can change into another. Last frames work with supported Kling choices, Veo, Wan, Seedance, and MiniMax. They are not available on the default Kling choice or Grok.

AI Video Models and Their Controls

The table shows the controls available in Secret Chat AI. A family may have more than one choice in the picker, so its range covers those choices. If you switch to a model that cannot serve the current length or resolution, the app steps down to the nearest supported value.

Model familyLengthResolutionText-to-video shapesSoundLast frame
Kling3–15 seconds, depending on the choice720p or 1080p9:16, 16:9, 1:1OptionalOn supported choices
Veo4, 6, or 8 seconds720p, 1080p, or 4K; 1080p and 4K use 8 seconds9:16, 16:9Always onYes
Grok1–15 seconds480p or 720p9:16, 16:9, 1:1, 4:3, 3:4, 3:2, 2:3OptionalNo
Wan2–30 seconds480p, 720p, or 1080p9:16, 16:9, 1:1, 4:3, 3:4OptionalYes
Seedance4–30 seconds, depending on the choice480p, 720p, or 1080p9:16, 16:9, 1:1, 4:3, 3:4, 21:9OptionalYes
MiniMax5–15 seconds480p or 768p9:16, 16:9, 1:1, 4:3, 3:4, 21:9No switch; a soundtrack may be returnedYes

Most models open at 5 seconds. Veo opens at 8 seconds. Sound is off by default where a switch is available. Pick only what the scene needs: a short test is useful for checking motion, while Wan and Seedance can carry an idea for up to 30 seconds.

How to Make a Clip

  1. Open Secret Chat AI and choose Video or Video Lax. Pick Video for Kling, Veo, or Grok. Pick Video Lax for Wan, Seedance, or MiniMax.
  2. Choose the model. Start with the length, resolution, aspect ratio, and sound controls it offers.
  3. Add frames if you want them. Use First frame for an AI video from photo. Add Last frame when the chosen model supports it.
  4. Write the action. Say what moves, how the camera behaves, and what should remain steady. Text is required when no first frame is attached.
  5. Check the displayed price and start. The full amount is held while the clip renders. A failure or refusal costs nothing.
  6. Download the result. The clip arrives in the chat as a video player with a Download button.

Settings Stay Close to the Message Box

An empty video chat opens with the full settings form. After the first request, the current choices sit in chips under the message box. The row covers the model, length, resolution, aspect ratio, sound, and the price. Frame pictures have fixed First frame and Last frame slots above the box; each picture can be up to 10 MB.

The last prompt and frame pictures are remembered in that browser. This makes it easy to change one setting and try again. The clip itself comes back into the conversation, and rendering usually takes from seconds to a few minutes. The app can keep waiting for up to 2 hours. Closing or reloading the page does not cancel the job.

The “I Can Wait” Video Setting

Kling, Wan, and Seedance show an hourglass chip called I can wait. It is on by default, so the app stays with the main provider. Turn it off if waiting matters more than avoiding a second render. When the first clip is still unfinished after 5 minutes, a backup provider starts the same job. Both clips may arrive, and you pay for each one that does. The backup clip is charged at the backup provider's own price, so the price shows “≈” while the switch is off.

The switch is not shown for Veo, Grok, or MiniMax. You can change it while a supported clip is still rendering.

Standard and Uncensored Video Generation

The Standard Video entry uses the stricter screening level. Video Lax uses the uncensored level, which gives lawful fiction, satire, dark themes, clinical work, political work, and other creative prompts more room. It means fewer needless refusals, while firm limits remain. Sexual content involving minors, terrorism, and real threats of violence are blocked on every model. Each provider also keeps its own rules, so a request can still be refused.

Read more about the lighter screening level on the uncensored AI video generator page or see the wider uncensored AI guide.

Where the Video Goes

A finished clip appears in the chat with playback controls and a Download button. It is also listed on the Library's Videos page. The library lives in browser storage on your device, and exports include your clips. Secret Chat AI keeps no cloud archive, so download or export work you do not want to lose.

This AI video maker uses a prepaid dollar balance. There is no subscription, and the balance never expires. Every model shows the clip price before you begin. Refusals, failed downloads, provider failures, and clips that time out cost nothing. For presentations, documents, social posts, and the other work tools, browse all generators.

Privacy, With the Limit Included

We store nothing on our side: there is no chat or video archive on Secret Chat AI's servers. History stays in your own browser. We build no profile of you and do not link a chat to you. Model providers see our service rather than the person making the request. Your material is never used for training.

Secret Chat AI removes you from your queries — it does not remove the data from your messages. Every prompt and frame picture goes verbatim to a third-party model provider, and the OpenAI classifier also sees the material during screening. A provider's retention rules may still apply. Remove names and any identifying or sensitive details before you send them.

Frequently Asked Questions

Can I make an AI video from a photo?
Yes. Put the photo in the First frame slot, describe the motion or scene, and start the clip. The video follows the photo's shape on most models; Veo keeps the aspect ratio you selected.
Can I generate video from text alone?
Yes. Leave the frame slots empty and write the scene you want. The selected model then creates a text-to-video clip in your chosen length, resolution, and aspect ratio.
Which model can make the longest clip?
Wan and the longer Seedance option can make clips up to 30 seconds. Other model families have shorter limits, which the settings show when you switch models.
Can an AI video include sound?
Yes, depending on the model. Kling, Grok, Wan, and Seedance have a sound switch. Veo always makes sound. MiniMax has no switch, though it may return a soundtrack.
Can I set both the first and last frame?
Yes, with supported Kling choices, Veo, Wan, Seedance, and MiniMax. Add the first picture before the last-frame slot becomes available. The default Kling choice and Grok do not accept a last frame.
What happens if a video generation fails?
A failed, refused, timed-out, or undeliverable clip costs nothing. The full displayed price is held while a clip renders and charged only when it is delivered. One exception: with I can wait turned off, if one clip arrives and the provider still charged for the other, unfinished clip, that charge is passed on at cost, with no markup.
Where is my generated video stored?
It is stored in your browser on your device and appears on the Library's Videos page. Secret Chat AI keeps no cloud archive of the clip, so download or export anything you want to keep.
What does the I can wait setting do?
When it is on, the app keeps waiting for the main provider. Turn it off and, if the clip is not ready in five minutes, a backup provider starts another render. Both clips can arrive, and each delivered clip is charged.