VEED Avatars

VEED Avatars — Text & Audio to Talking Avatar Video

What is VEED Avatars?

VEED Avatars is a UGC-style talking avatar video API built by VEED and served through Segmind. Give it a script or an audio clip, choose one of 28 stock presenters, and it returns a polished MP4 of that avatar speaking your words with realistic lip-sync, facial expressions, and natural body movement. It is designed for creators, marketers, and developers who need on-brand talking-head video at scale without cameras, actors, studios, or re-shoots.

The model runs in two modes. In text mode, VEED's built-in text-to-speech voices your script automatically, so you only provide words. In audio mode, you supply your own speech recording and the avatar's mouth, head, and gestures are driven directly by that audio, giving you full control over voice, tone, and language. You pick a presenter with the avatar_id dropdown — options span named avatars such as Emily, Marcus, Mira, Jasmine, Aisha, and Elena, plus generic Any Male and Any Female presenters, each available in vertical, standard, side, and walking framings.

Key Features

• •28 ready-to-use stock avatars across vertical (9:16) and standard orientations, including walking and side poses.
• •Two input modes: text-to-speech from a script, or audio-driven lip-sync from your own recording.
• •Realistic, synchronized lip-sync with natural facial expressions and body motion.
• •Synchronous API — POST your request and receive the finished MP4 directly, no polling.
• •Per-minute pricing so cost scales with output length.

Best Use Cases

VEED Avatars fits any workflow that needs a person on camera without filming one. It excels at short-form UGC ads for TikTok, Instagram Reels, and YouTube Shorts, where the vertical avatars deliver native 9:16 framing. Use it for product explainers and demos, onboarding and training clips, personalized outreach and sales videos, faceless-channel content, and localized marketing where the same script is regenerated instead of re-shot. Standard and side framings suit landscape ad placements and website hero videos, while walking avatars add dynamism to lifestyle and brand spots.

Prompt Tips and Output Quality

Provide exactly one of text or audio_url — never both. For the cleanest results in text mode, write conversational, punctuation-rich scripts and keep sentences short so the TTS phrasing lands naturally. For precise voice, accent, or non-English delivery, record or synthesize audio yourself and pass it via audio_url. Match the avatar framing to your placement: pick a _vertical_ avatar for Reels and Shorts, and a standard or _side_ avatar for landscape. Because billing is per minute of output, trim your script to only the words you need.

FAQs

What is the difference between text mode and audio mode? Text mode voices your script with VEED's built-in TTS. Audio mode uses a speech file you supply to drive the avatar. Provide only one.

How many avatars are available? 28 stock avatars, including Emily, Marcus, Mira, Jasmine, Aisha, and Elena, plus generic Any Male and Any Female presenters, in vertical, standard, side, and walking variants.

Can I make vertical videos for TikTok and Reels? Yes. Choose any avatar_id containing vertical to get native 9:16 output.

What format is the output? The API returns an MP4 video file directly in a synchronous response.

Can I use my own voice? Yes. Pass a speech recording to audio_url and the avatar will lip-sync and gesture to it.

Is it suitable for commercial and UGC ads? Yes. VEED Avatars is built for UGC-style ads, product explainers, and social content, and permits commercial use.