← Blog

AI videoNovember 18, 2026

How to make a talking pet video

Make your dog or cat talk with AI: one photo, one short line, one video model. Two ways to do it, prompts to copy, common fixes, and real prices from $1.35.

To make a talking pet video, give an image-to-video model a clear, front-facing photo of your pet and the line you want it to say, with sound turned on. For exact words in a voice you pick, make the voice first and add lip sync. On Vaaya the quick way costs $1.35 a clip.

What you get

A 5 to 10 second clip of your pet looking at the camera and "saying" a short line. The mouth moves, the head tilts, and you hear a voice. Popular formats: the dog explaining why the couch is wet, the cat filing a complaint about breakfast, the pet reviewing a new toy.

What you need

  • A photo of your own pet, or one the owner said you can use. Facing the camera, eyes and mouth visible, good light. Eye level works best.
  • One short line, under 15 words.
  • An idea of the voice: deep and slow, squeaky, posh, grumpy.
  • An AI agent connected to Vaaya, or another way to run an image-to-video model.

Two ways to do it

Quick: let the video model make the voice. Some image-to-video models make sound along with the picture. Put the line in quotes in the prompt, describe the voice, and turn audio on. You get picture and voice from one call. The exact words can drift, so keep the line short.

Exact: make the voice first, then lip sync. Write the line and turn it into speech with a voice you pick. Animate the photo. Then run a lip sync model to match the mouth to your audio. This costs more. Lip sync models are trained mostly on human faces, so test one clip of your pet before you make many.

Step by step (quick way)

  1. Crop the photo so your pet fills most of the frame. Use vertical 9:16 for Reels or TikTok.
  2. Write the line and the voice. For example: a deep, tired voice says "I did not eat the sandwich. The sandwich left."
  3. Generate a 5-second clip with audio on.
  4. Watch it with sound. Check that the mouth starts moving when the voice starts.
  5. If the words come out wrong, shorten the line, or switch to the exact way.

Prompts you can copy

Dog:

The dog looks straight at the camera and talks, jaw moving in time with
the words. A deep, slow, tired male voice says: "I did not eat the
sandwich. The sandwich left." The dog tilts its head at the end. Camera
still. Keep the dog's face, fur and markings the same as the photo.
No music.

Cat:

The cat stares at the camera and speaks with small, precise mouth
movements. A posh, unimpressed female voice says: "Breakfast was four
minutes late. I will be speaking to the manager." Slow blink at the end.
Camera still. Keep the cat's face and fur the same. No music.

Puppy-energy version:

The dog bounces slightly and talks fast, tail wagging at the edge of the
frame. A high, excited voice says: "Is that a walk? Are we going? We are
going!" Camera still. Keep the dog's face and markings the same.

Common mistakes and fixes

Mistake Fix
Side-on photo Use a photo where both eyes face the camera. The model needs to see the mouth.
A long speech One or two short sentences per clip. Join clips if you need more.
Your pet turns into a cartoon Add "realistic, keep the fur and markings the same as the photo."
Human-looking lips appear Ask for "jaw movement only" or "small mouth movements."
The voice does not suit the pet Describe age, pace and mood of the voice, not just "a funny voice."
Posting someone else's pet Ask the owner first, and say the video is AI-made.

Make it on Vaaya

There is no talking-pet recipe on Vaaya. You run it through an AI agent connected to Vaaya, such as Claude Code, Cursor or Codex. Vaaya is a wallet for your agents: you load one balance, and the agent sees the price before each call.

The tools, with their catalog prices in November 2026:

Step Tool Price
Upload the photo fal/upload 1¢
Quick way: clip with voice seedance-2-0--fast--image-to-video on fal/generate, audio on $1.35
Exact way: voice line deepgram/speak 1¢ per 250 characters, 2¢ minimum
Exact way: lip sync sync-lipsync--v2 on fal/generate $5.00

The quick way is $1.36 with the upload. The exact way is about $6.40: the voice line, the $1.35 clip with audio on, then $5.00 for lip sync. ElevenLabs voices are also on fal/generate as elevenlabs--tts--turbo-v2-5, at about 5¢ per 1,000 characters.

Connect your agent to Vaaya, then paste this:

Use Vaaya to make a talking pet video from [photo path].
Upload the photo with fal/upload. Then call fal/generate with model
seedance-2-0--fast--image-to-video, image_url set to the upload,
generate_audio true, duration 5, resolution 720p, and this prompt:
[paste your prompt].
Tell me the price first and stay under $3. Poll the job with result
instead of submitting it again. Save the MP4.

New to animating photos? Start with how to turn a photo into a video with your agent.

Questions

How do I make my pet talk in a video?

Give an image-to-video model a clear, front-facing photo of your pet and a prompt with the line in quotes, with sound turned on. For exact words in a voice you choose, make the voice with text-to-speech first and add lip sync.

How much does a talking pet video cost?

On Vaaya, the quick way is $1.35 for a 5-second clip plus 1 cent to upload the photo. The exact way, with your own voice line and lip sync, is about $6.40.

Why does my pet say the wrong words?

When the video model makes the voice, long lines drift. Keep it under 15 words, put the line in quotes, and describe the voice. For exact wording, use text-to-speech and lip sync.

What kind of photo works best?

Your pet facing the camera at eye level, both eyes and the mouth visible, good light, and a plain background. Side views and blurry photos give the model too little to work with.

Make your pet talk with your agent

Connect Claude Code, Cursor or another agent to Vaaya. Load one balance and see the price before every call.

Connect your agent