Skip to main content

Talking photo AI

Pick a portrait photo. After you sign in, attach it with an audio track, or a script for a workspace voice to read, and Endless turns it into a talking video with MiniMax H3 Max Lip-sync.

Video
Turn the photo I attached, host-portrait.jpg, into a talking video with MiniMax H3 Max Lip-sync. I will attach the audio, or give you a script to read in one of the workspace voices. Tell me the cost before you run it.
Photo · host-portrait.jpg
Audio · episode-12-intro.mp3, 42 seconds
Model · MiniMax H3 Max Lip-sync, 720p

Here is a 42-second talking video from the portrait. The mouth follows the episode intro from start to finish.

  • 42 seconds, the length of the audio.
  • 720p. Ask for 1080p or 2K for a sharper take.

What is a talking photo?

A talking photo is a video made from one still portrait and an audio track. The model animates the face so the mouth follows the speech, and the video lasts as long as the audio. No filming is needed, only a clear photo and the words you want said. The photo can be of a real person or a character you made.

People use talking photos for presenter videos, course intros, product explainers, greetings and characters in social posts. It turns a headshot or a generated character into a speaking video in minutes. The result works best with a front-facing portrait where the face is well lit and clearly visible.

Endless runs MiniMax H3 Max Lip-sync in your workspace chat. You attach the photo and an audio file between 5 seconds and 15 minutes long, or give the agent a script and a workspace voice reads it first. The model needs no prompt. You can pick 480p, 720p, 1080p or 2K, and the video is saved in your workspace.

How it works

  1. 01

    Pick the photo

    Choose a portrait in the box above. Only its name goes into the prompt; the file is not uploaded from this page.

  2. 02

    Sign in and attach the files

    Create a free Endless account or sign in. Your workspace chat opens with the prompt already written. Attach the same photo and the audio, or paste a script for a voice to read, and send.

  3. 03

    Get the talking video

    The agent runs MiniMax H3 Max Lip-sync and saves the video in your workspace. Ask for a higher resolution, a new script or a different voice.

What people use it for

  1. 01

    Presenter videos without filming

    Turn a headshot into a presenter who reads your script for a course, an onboarding video or an internal update. When the script changes, make a new take from the same photo instead of filming again.

  2. 02

    Talking characters

    Generate a character with an image model in Endless, then give it a voice and a line. The image and the talking video stay in the same workspace, so you can reuse the character across posts.

  3. 03

    Product explainers

    Have a spokesperson photo explain a feature, a launch or an offer in a short clip for a product page or a social post. A short, clear script tends to work better than a long one.

  4. 04

    Personal greetings

    Make a welcome or thank-you message from your own photo and your own voice. Record the message and attach it, or clone your voice from recordings of you and have the agent read a script.

  5. 05

    Videos in other languages

    Ask the agent to translate the script, have a workspace voice read it, and make a version of the same talking photo for each language. Each version is saved in your workspace for your team.

  6. 06

    Long narration

    The model accepts audio of up to 15 minutes, so a portrait can narrate a full lesson, a podcast segment or a walkthrough. Longer audio costs more, so ask the agent for the cost first.

Frequently asked questions

  • Is talking photo AI free?

    You can start free. A free Endless account comes with the Free plan, which gives your workspace 100 credits when you create it and 50 credits every week. MiniMax H3 Max Lip-sync is available on the Free plan, and each run spends credits from that balance. Paid plans add more credits.

  • How much does a talking photo cost?

    It charges per second of video, and the video is as long as the audio, so longer audio costs more. Higher resolutions cost more too. You can ask the agent for the cost before it runs, so you know what it will spend from your workspace balance.

  • What photo works best?

    A front-facing portrait with the face well lit and clearly visible. Avoid hands or objects over the mouth and faces that are very small in the frame. Use the highest-quality version of the photo you have.

  • How long can the audio be?

    Between 5 seconds and 15 minutes, and the video lasts exactly as long as the audio. If you give a script instead, a workspace voice reads it first and the length follows the reading.

  • Can I use a photo of someone else?

    Only with their permission. Use your own photo, a photo of someone who agreed to it, or a character you generated, and do not make a real person appear to say something they did not agree to say. Endless asks you to confirm consent before it clones a voice.

  • What is the best talking photo AI?

    Look for natural mouth movement, support for long audio, a choice of resolution, and a way to use your own audio or a generated voice. Endless fits because it runs MiniMax H3 Max Lip-sync next to voices and image models, with every result saved as an asset.

  • Is my photo private?

    Photos and audio you attach, and the videos made from them, are saved in your workspace. They are private by default, and you choose whether to share them with teammates. Endless does not post or publish your files anywhere.

Related tools

Give your next idea a place to grow.

Start free, create a project, and bring your brand’s references together. Or book a demo to explore Endless with your team.

Book a demo