New in 4.0 Talking hosts and stick-figure scenes

A cartoon host who talks with your voice.

Put a host beside your pictures. Its lips follow every sound of the narration, it points at what is shown, shrugs at a question and reacts to the story, and the AI Director decides where it appears. Add stick-figure scenes drawn from your script, in the explainer-channel style.

  • Any voice, including your own recording
  • 20 hosts, or build your own
  • Lip sync on your computer, no account
  • 9:16 and 16:9

Ten thousand years in five minutes Exported from the app, 3:34. The History narrator is placed and directed by the AI Director, with stick-figure scenes, charts and a full-frame question.

  • 20ready-made hosts
  • 12mouth shapes
  • 5moods per sentence
  • 145drawn characters and objects
  • 0credits for the host
How it works

Pick a host, and the Director does the rest

The host is planned with the pictures, so it knows the story it is telling.

  1. 1

    Pick a host

    Choose a Channel host in the wizard, or start from a template that has one. Or press Host on any scene of the Visuals screen, or any board in Doodle Maker.

  2. 2

    The Director directs it

    In the same plan that picks the pictures, the Director decides which scenes the host is in, where it stands, and the gesture and mood of every sentence.

  3. 3

    Change anything

    Change the gesture and mood of any sentence, move the host, or switch it on or off on every scene at once.

  4. 4

    Preview and export

    The host plays in the live preview as you scrub, exactly as it will be exported.

Samples

Clips from real exports

Every clip with a voice is cut from a video exported by the app, unchanged. Press play to hear the lips follow the voice.

A host in the corner

The History narrator waves hello and presents the first figure, small in the corner like a news anchor.

Full frame for the big question

The host steps up to fill the frame to ask the viewer a question, then goes back beside the pictures.

Your own character

A host made on the Hosts screen, with the channel name on its shirt, standing beside the picture.

"Out of nine workers, four..."

A stick-figure scene: a row of nine, with four turning red on "four", a place tag and a handwritten label.

The inflatable decoy tank

Four soldiers and a drawn decoy tank that deflates, each brought in on the word that names it.

People and things, on their words

A builder, a lorry, a nurse and a hospital, with a label arrowed to what it names.

A rocket, a satellite, 1969

Drawn objects from the library with a date tag, while the host presents from the corner.

Pointing label No voice

New text on picture: a handwritten note writes itself on with a curved arrow to the spot you choose.

Lip sync and acting

It speaks the way your narrator speaks

The host works out what to do from your words, sentence by sentence.

The History narrator host in six poses: calm, pointing and happy, shrugging and worried, explaining, waving, and presenting
Calm · Point · Shrug · Explain · Wave · Present

Lips that follow the sounds

12 mouth shapes, from the actual sounds being said: closed for m, b and p, wide open for "ah", round for "oh", the lip under the teeth for f and v. Timed so they never flicker.

Gestures as it speaks

It points at what is being shown, shrugs at a question, explains with both hands, presents, and waves hello at the start of the video.

A mood for every sentence

Its brows and mouth show calm, happy, worried, angry or sad. Brows lift and the head nods on the words it stresses.

Never frozen

It blinks, sways and nods between words, and a host that carries on into the next scene joins it without a break.

Any voice

It works with any voice in the app, and with your own recording. The lip sync runs on your computer and needs no account.

Always faces the picture

Stand it on the left, right or centre, and it turns to the picture, so it points the right way from either side.

Directed by the AI Director

On screen when it helps, out of the way when it doesn't

Pick a Channel host in the wizard, or start from a template that has one: News explainer brings a news anchor in the corner, History a history narrator, Shorts explainer a teacher.

It opens and closes the video

The host greets the viewer, asks the questions, turns the story, presents the charts and maps, and signs off.

It steps aside

Where a picture should speak alone, the host is off. Beside the pictures it is on in at most about six scenes in ten, and never gone for more than three in a row.

It acts on your words

It points at a figure, shrugs at a question, looks worried at bad news and sad at a loss.

It stands where it fits

With The Director picks, it stands on the side away from what the picture shows and any title on it, and keeps a side for a while instead of hopping.

It never covers a figure

Over a chart, map or timeline the host moves to the corner, and it goes full frame only for a big question to the viewer.

Your choices stay yours

A host you set on a scene yourself is never changed by planning again. Pick a place of your own and it keeps to it all video.

Where it stands is yours to set. Left, Right, Centre, Corner, like a news anchor, Full frame, or The Director picks, scene by scene.

20 hosts

A host for every kind of channel

Every one is the app's own design, free to use in paid videos. Open More characters in the host picker for the 75 characters of the drawn library too.

The 20 ready-made hosts, from a news anchor to a cartoon dog
  • News anchor
  • Soldier
  • Scientist
  • Detective
  • Teacher
  • Robot
  • Builder
  • Stick person
  • History narrator
  • Explorer
  • Pilot
  • Doctor
  • Chef
  • Farmer
  • Businessperson
  • Police officer
  • Student
  • Parent
  • Shopkeeper
  • Cartoon dog
Make your own

Build your channel's own host

The Hosts screen in the sidebar builds a character from the same parts as the ready-made ones. Start from any of them, and a live preview says a line as you build, in any gesture and mood. Characters are saved as you change them.

Face

A person, robot or animal face, with the skin or fur you choose.

Hair and hats

17 hairstyles and hats, or none, in any colour. Crowns, helmets, a pharaoh's headdress, a turban and more from the library.

Outfit

17 outfits with a main and a second colour, plus robes, armour, a toga, a dress and a spacesuit.

Extras

Glasses, a moustache, a beard, a bow tie, a stethoscope, a bandolier.

Your name on the shirt

Put your channel name or a short word on the shirt.

Share it

Export a character as a small .json file and import it on another computer.

Your videos stay the same. A video keeps its own copy of the character, so it renders the same on another computer and after you change or delete the character. When you change it, the Host panel offers Update this video.

Stick-figure scenes

Explainer scenes, drawn from your script

Tick Stick-figure scenes in the Director's footage and it draws simple scenes on paper where a line needs one. Each thing comes in as it is said: the soldiers on "soldiers", the tank on "tank", the red figures of a row on their number.

Counting people

"4 of 9": a row of cartoon people, with the ones the line singles out turning red and worried.

"Imagine you are..."

The person the line asks you to be, in their own place and time.

Two sides

Two sides set against each other, each with the line's own word over it.

Cause and effect

A cause and what it leads to, with handwritten labels that write themselves on with an arrow.

The 75 characters of the drawn library, from cavepeople and pharaohs to astronauts, a wizard and a zombie
75 characters from every era, and 70 objects that move: tanks drive and fire, rockets launch, flags flutter, clouds rain, volcanoes erupt, dinosaurs walk and roar.

Something the library does not have? It is shown as a line icon from a set of nearly a thousand, or, when no icon fits, as one line drawing made with your image provider. Only a thing the narration says is drawn this way, at most one picture a scene, and a label never shows a number the narration does not say.

FAQ

Questions about talking hosts

What it does, what it works with and what it costs.

Still have a question? Send us a message and we'll help you out. Contact us
What is a talking host?

A cartoon character that stands beside your pictures and talks with your video's voice. Its lips follow the sounds being said, and it points, shrugs, explains and reacts with a mood for each sentence. It is new in Vid Optimus 4.0, for AI videos and Doodle Maker.

Is it a realistic AI avatar?

No. It is a cartoon host, drawn by the app, in the style of faceless explainer channels. It is not made from a photo or a video of a real person.

Which voices does it work with?

Any voice in the app, and your own recording. The lip sync runs on your computer and needs no account. Its mouth shapes are learned from English speech, so English narration gets the most exact lips.

Does the host cost anything to use?

No. There are no credits or per-video fees for the host. The lip sync runs on your computer, and every character is the app's own design, free to use in paid videos.

Can I make my own host?

Yes. The Hosts screen builds a character from the same parts as the ready-made ones: a person, robot or animal face, hair or a hat, an outfit in your colours, extras, and your channel name on the shirt. A live preview says a line as you build. Make it your Channel host and every new video starts with it.

Do I have to place the host on every scene?

No. Pick a Channel host in the wizard and the AI Director decides which scenes it appears in, where it stands, and the gesture and mood of every sentence. Or choose Every scene to keep it on the whole video.

Can I change what the host does?

Yes. Change the gesture and mood of any sentence, move it, or turn it off. Use this host on every scene gives the whole video the same host, and Switch on everywhere or Switch off everywhere does every scene at once.

Will the host cover my captions or charts?

No. It sits over the picture and under the captions. Over a chart, map or timeline it moves to the corner, and in a tall video it stands a little taller when the captions would cross its mouth.

What are stick-figure scenes?

Simple drawn scenes made from your script in the explainer-channel style: rows of cartoon people, drawn objects that move, a place or date tag and handwritten labels. Tick Stick-figure scenes in the Director's footage and it draws them where a line counts people, asks you to imagine being someone, sets two sides against each other or shows a cause and its effect.

What can stick-figure scenes draw?

75 characters from every era, from cavepeople to astronauts, and 70 objects, from tanks and steam trains to castles, money and dinosaurs, all the app's own drawings. When a line names something the library does not have, it is shown as a line icon, or as one line drawing made with your image provider. At most one picture is made for a scene.

Does it work in Doodle Maker?

Yes. Pick a host on any board of Doodle Maker's Boards step, or as the Channel host in its wizard, and the boards are drawn beside the host rather than under it.

Does it work in vertical videos?

Yes, in 9:16 and 16:9. In a tall video the stick-figure scene rises above a host standing across the bottom.

Get started

Give your faceless channel a face.

Free for 7 days. No card needed. Works on Windows and macOS.