Atlas Cloud is now an authorized U.S. reseller of Dreamina Seedance 2.5 and Seedream 5.0 Pro.
Avatar Omni Human 2.0 Multi-Person Avatars

Avatar Omni Human 2.0 Multi-Person Avatars

ByteDance's Avatar Omni Human 2.0 is an audio-driven model for character animation. Support for multi-person scenes and dynamic gestures extends avatar creation to shared performances with expressive character movement. On Atlas Cloud, integrate the model through API access and use its model-specific parameter schema to configure inputs for your video workflow. Start building today.

Avatar Omni Human 2.0 is developed by ByteDance. Atlas Cloud (operated by Atlas Cloud AI LLC) provides access to it and does not own it. All trademarks belong to their respective owners.

Avatar Omni Human 2.0 Input and Video Output Specs

Compare the image and audio inputs, video output, and verified capabilities of the Avatar Omni Human 2.0 endpoint.

ModalityDescription
Avatar Omni Human 2.0 API (Image and Audio To Video)Turn one image and an audio track into a 720p video of one or more characters speaking or singing. Avatar Omni Human 2.0 supports dynamic gestures, multi-person scenes, prompt-directed camera movement, and clips up to five minutes, making it suitable for avatar presentations, performances, and dialogue videos.

Avatar Omni Human 2.0: From Voice to Performance

Avatar Omni Human 2.0 turns one image and an audio track into 720p speaking or singing videos with dynamic gestures, scenes with multiple characters, prompt directed camera movement, clips up to five minutes, and unified pay as you go API access through Atlas Cloud.

Avatar Omni Human 2.0 Image to Performance

Start with one image and one audio track to create a 720p video of the character speaking or singing. Avatar Omni Human 2.0 uses the source image to establish the character while visible movement and dynamic gestures shape the performance. This compact workflow turns a still character into presenter clips, music visuals, and narrative moments while keeping image and audio as the core inputs.

Audio Driven Performance

Audio drives the speaking or singing performance, shaping timing and dynamic gestures throughout the clip. Spoken lines and sung phrases provide the rhythm while the reference image establishes the character on screen. The result fits narration, musical performances, and dialogue scenes where visible delivery needs to feel connected to the soundtrack from the first beat to the last.

Five Minute Character Performances

Need room for a full explanation or song? Avatar Omni Human 2.0 can generate clips up to five minutes from one image and an audio track, extending beyond short social snippets. Use that duration for lessons, product walkthroughs, performances, or serialized character content. Longer outputs reduce the need to divide a single performance into many separate videos.

Prompt Directed Camera Stories

Guide the performance with a text prompt that can direct camera movement around the character and scene. Shift from a close view to a wider composition, request an orbit, or shape how the framing follows the audio driven performance. This added direction gives creators more control over explainers, narrative moments, and music driven sequences while preserving the simple image and audio workflow.

Avatar Omni Human 2.0 Ensemble Motion

Bring several characters into the same Avatar Omni Human 2.0 scene for shared speaking or singing performances. Dynamic gestures and audio driven motion animate the cast within one composition rather than isolating every character. Interviews, duets, ensemble performances, and scripted exchanges gain a more natural sense of interaction when the story depends on more than one visible participant.

Atlas Cloud Access for Avatar Omni Human 2.0

Connect Avatar Omni Human 2.0 to an existing product through Atlas Cloud's unified API and one OpenAI compatible key. Transparent pay as you go pricing lets developers generate only the video volume they need, while the same integration reaches other Atlas Cloud models. This setup suits applications and automated content pipelines that need avatar generation without managing a separate integration for every model.

Avatar Omni Human 2.0 in a One Prompt Model Test

Compare how Avatar Omni Human 2.0 and two alternative video models interpret identical prompts across realistic performance and stylized action scenes.

Prompt

Create a 9-second cinematic action sequence that faithfully preserves the reference portrait’s facial identity, transforming the subject into a 1960s retro-science-fiction alpine observatory mechanic. In a cobalt-blue blizzard, the runaway dome suddenly spins as exposed brass gears clash and spark; begin with a rapid low-angle push across the snowy ground as the mechanic lunges onto frost-covered rails, grabs a wildly swinging brass telescope, and struggles forward, his scarlet scarf and work coat snapping violently in the wind. Cut to a tight orbiting tracking shot as he ducks beneath a sweeping ladder frame, braces against the telescope’s momentum, and reaches for the brake lever amid swirling snow mist and colliding machinery. Cut to a dramatic overhead view inside the rotating dome: a crack in the roof briefly aligns with the luminous Milky Way, bathing his consistent face in starlight as he yanks the brake; the gears shudder, the dome stops, and the telescope settles with believable weight. Whip-pan to his nearby steaming coffee, then crash-zoom into the cup as a tiny glowing meteorite plops into it with a sharp hiss; cut to his close-up as startled disbelief becomes a barely suppressed grin. Continuous energetic motion, seamless stunt choreography, realistic cloth, snow, inertia, impacts, sparks, and mechanical physics; no slow motion. Technicolor 1960s adventure-film palette, exquisitely detailed practical miniature-model craftsmanship, cobalt-blue snow night, burnished copper-gold machinery, vivid scarlet scarf, dramatic hard backlight, subtle film grain, anamorphic flares, tactile analog effects. Audio: roaring wind, metallic gear crashes, rattling rails, straining cables, a decisive brake clank, sudden mechanical silence, then the meteorite’s comic plop and sizzle, punctuated by a brisk retro orchestral sci-fi score. 16:9 aspect ratio.

Generated with Avatar Omni Human 2.0 on Atlas Cloud

Generated with Gemini Omni Flash Image-to-Video on Atlas Cloud

Generated with Avatar Omni Human 1.5 on Atlas Cloud

Prompt

A 9-second cinematic micro-story: the same photorealistic courier in a vivid vermilion raincoat, with consistent face, clothing, and proportions throughout, accidentally enters a colossal handcrafted origami city just after a torrential downpour. In a cold cyan-blue alley of towering folded-paper buildings, warm amber windowlight glowing through their creases, a giant paper crane snatches the courier’s parcel and darts away. Start with an ultra-low ground-level tracking shot racing beside the courier’s pounding boots as they sprint through mirror-like puddles, reflections forming striking symmetrical compositions; raindrops splash toward the lens and massive folded-paper foreground shapes sweep past in strong perspective. The courier vaults a puddle and ducks beneath a curling paper bridge that flexes, wrinkles, and crackles with believable paper physics. Whip-pan seamlessly into the crane’s top-down aerial POV as it banks between origami rooftops, the courier still running below; then dive rapidly toward them and transition into a tight 360-degree orbit as the courier leaps, catches the crane, and snatches back the parcel midair. The instant they land, the parcel playfully unfolds itself with precise stop-motion creases into an enormous paper umbrella; its canopy snaps open and lifts both the startled courier and flapping crane into the rainy sky for a whimsical final reveal. Hybrid aesthetic combining a lifelike human performance, cinematic rain and reflections with tactile handmade origami stop-motion texture; crisp paper fibers, wet surfaces, convincing weight, momentum, splashes, bending folds, and continuous complex motion. Layered framing through foreground paper edges, dramatic wide-angle distortion, cool teal storm light anchored by the red coat and warm yellow windows. Rhythmic synchronized audio: raindrops, wet footsteps, sharp paper crackles, wing flaps, and the umbrella’s final percussive pop perfectly aligned with each movement; playful minimal percussion builds through the chase and lands on the airborne twist. No screens, interfaces, dashboards, progress bars, charts, captions, logos, subtitles, readable text, frozen poses, slow-motion filler, identity drift, costume changes, duplicate limbs, broken anatomy, teleportation, discontinuous motion, incoherent cuts, plastic-looking paper, or implausible physics. 16:9 aspect ratio.

Generated with Avatar Omni Human 2.0 on Atlas Cloud

Generated with Gemini Omni Flash Image-to-Video on Atlas Cloud

Generated with Avatar Omni Human 1.5 on Atlas Cloud

Direct Every Performance with Avatar Omni Human 2.0

From product explainers and lessons to multi-person dialogue, singing performances, social stories, and long form updates, Avatar Omni Human 2.0 turns one image plus audio into prompt-directed 720p avatar video.

Product Demos with Avatar Omni Human 2.0

Turn one character image and a narration track into a 720p presenter video with prompt-directed camera movement. Product teams can produce guided demos, launch explainers, and feature walkthroughs without filming a live host.

Courses and Team Training

Create five-minute teaching videos from a single image and recorded lesson audio, supported by dynamic gestures. Educators and training teams can publish lectures, onboarding modules, and policy briefings with a consistent on-screen presenter.

Avatar Omni Human 2.0 Dialogue Scenes

Stage conversations from one image and an audio track while Avatar Omni Human 2.0 animates multiple people. Media creators can build interview segments, panel discussions, or character exchanges with coordinated speaking performances.

Singing with Avatar Omni Human 2.0

Pair a character image with music or vocals to generate a 720p singing performance with expressive gestures. Artists and entertainment teams can create virtual performances, music promos, and character-led fan content.

Directed Social Stories

Direct camera movement and character behavior through prompts, then let the supplied audio drive the performance. Social teams gain reusable talking-character clips for announcements, scripted stories, campaign posts, and episodic content.

Long Form Leadership Updates

Keep one avatar speaking or singing for clips up to five minutes, with gestures that follow the audio. Organizations can publish executive updates, event messages, narrated explainers, and recurring communications in a consistent format.

Avatar Omni Human 2.0 Across Leading Video Models

Compare Avatar Omni Human 2.0 with three Atlas Cloud video models by provider, input workflow, clip length, resolution, and catalog base price.

ModelProviderInput WorkflowMax Clip LengthOutput ResolutionCatalog Base Price
Avatar Omni Human 2.0ByteDanceImage + audio + optional promptUp to 5 minutes720p$0.07
Avatar Omni Human 1.5ByteDanceImage + audio + optional promptUp to 60 seconds at 720p720p or 1080p$0.12
MiniMax H3 Image-to-VideoMiniMaxFirst frame + optional last frame + prompt5 to 15 secondsUp to 2K$0.038 per second
Veo 3.1 Lite Start-End Frame to VideoGoogleFirst frame + last frame + prompt8 seconds720p or 1080p$0.05

How to Use Avatar Omni Human 2.0 on Atlas Cloud

Get started in minutes — follow these simple steps to integrate and deploy models through Atlas Cloud's platform.

Create an Atlas Cloud Account

Sign up at atlascloud.ai and complete verification. New users receive free credits to explore the platform and test models.

Why Use Avatar Omni Human 2.0 on Atlas Cloud

Combining the advanced Avatar Omni Human 2.0 models with Atlas Cloud's GPU-accelerated platform provides unmatched performance, scalability, and developer experience.

Performance & flexibility

Low Latency:
GPU-optimized inference for real-time reasoning.

Unified API:
Run Avatar Omni Human 2.0, GPT, Gemini, and DeepSeek with one integration.

Transparent Pricing:
Predictable per-token billing with serverless options.

Enterprise & Scale

Developer Experience:
SDKs, analytics, fine-tuning tools, and templates.

Reliability:
99.99% uptime, RBAC, and compliance-ready logging.

Security & Compliance:
SOC 2 Type II, HIPAA alignment, data sovereignty in US.

Avatar Omni Human 2.0 Developer FAQ

Avatar Omni Human 2.0 is ByteDance's audio driven avatar video model. It transforms one image and an audio track into a 720p video of the character speaking or singing with synchronized expressions and gestures.

Create talking presenters, digital singers, character performances, and multi person scenes from a single reference image. Optional prompts can shape actions, scene elements, and camera movement for more directed results.

Create an Atlas Cloud API key and select bytedance/avatar-omni-human-v2.0 as the model in your request. Provide one image and one audio track, then include an optional prompt when you need additional performance or camera control.

Each request combines one reference image with an audio track. You may also supply a text prompt to guide the scene, character movement, specific actions, and camera behavior.

Yes, the model can animate characters performing speech or songs and supports scenes containing multiple people. Audio drives the performance while gestures and reactions are generated within the selected composition.

Videos are generated at 720p and can run for up to five minutes. This duration supports longer presentations, performances, and narrated character content without splitting the project into several short clips.

Atlas Cloud lists a standard price of $0.07 per second. Usage follows pay as you go pricing, so the total cost scales with the generated video duration.

Basic lip sync tools primarily coordinate mouth movement with speech. This model also aligns expressions and gestures with the audio while accepting prompt guidance for actions, scene details, and camera movement.

First confirm that the request includes both an image and an audio track and that any referenced media URLs are accessible. Compare every field with the current Atlas Cloud playground schema, then use the returned error details to correct the payload before retrying.

Explore More Families

Seedance 2.5

Seedance 2.5 API is now available on Atlas Cloud! It gives developers ByteDance's newest video model. It generates up to 30 seconds of native video in a single pass from text, a single image, or as many as 50 multimodal references, with synchronized audio and in-frame multilingual text. On Atlas Cloud you reach it through one key, with subject consistency and improved physics keeping long shots coherent. (Update: Seedance 2.5 1080P API Is Available NOW!)

View Family

Wan 3.0

Wan 3.0 API is the next generation of Alibaba's Wan video family, built to push long-form generation, multi-reference control, and audiovisual quality to new heights. Atlas Cloud already hosts Wan 2.7, 2.6, and 2.5, and Wan 3.0 runs on the same unified key with no separate setup. Start building today. Scroll down to the showcase to see what Wan 3.0 can create.

View Family

MiniMax H3

MiniMax H3 is MiniMax's multimodal video family for text, image, and reference guided creation. Across supported routes, it preserves subjects from reference media, offers flexible aspect ratios, and pairs generated sound with visuals through H3 Developer, with output profiles selected by endpoint. Atlas Cloud unifies the family behind one OpenAI-compatible key with transparent pay-as-you-go pricing from the standard rate of $0.038 per second. Start building today.

View Family

GPT Image 2.5

The gpt-image-2.5 family from OpenAI gives developers a choice of Flare and Sunburst for production image workflows. Render at arbitrary resolutions up to 3840x2160 and select from five quality tiers, including xhigh and max, to match specific output requirements. Atlas Cloud provides ready-to-use REST inference with no cold starts and standard pricing from $0.004 per generation. Start building today.

View Family

Seedream 5.0 Pro

Seedream 5.0 Pro API gives developers ByteDance's controllable image editing model on Atlas Cloud. It places edits precisely with anchors and coordinates, separates images into editable layers, fuses multiple references, and matches exact colors and materials, with multilingual text at 2K and 3K. On Atlas Cloud you reach it through one key!

View Family

Seedance 2.0

Seedance 2.0 is ByteDance’s production video model for precise shot creation. Turn prompts into video, animate a first-frame image with optional last-frame guidance, or shape results with reference media and optional web search. Atlas Cloud brings these workflows into one unified API with transparent pay-as-you-go pricing and one OpenAI-compatible key. Start building today.

View Family

Kling V3.0

The Kling 3.0 API brings Kuaishou's flagship video suite to Atlas Cloud through one OpenAI-compatible key. It spans two models, Kling 3.0 for AI Director storytelling, multilingual lip-sync, and precise on-screen text, and Kling 3.0 Omni (O3) for subject and voice cloning from a short video or image. Both generate native audio in the same pass, with output up to 4K. Build cinematic narratives, global marketing, multilingual ads, and serialized character content on reliable infrastructure.

View Family

Grok Imagine Video

Grok Imagine Video gives developers access to xAI’s video family for creating, animating, extending, and editing clips. Generate 1 to 15 second videos from prompts, guide people, objects, or styles with up to seven reference images, and work at 480p or 720p. Atlas Cloud brings these capabilities together through OpenAI-compatible endpoints with transparent pay-as-you-go pricing. Start building today.

View Family

Vidu Q4 Preview

Vidu Q4 is Vidu's latest flagship video model, designed for image-to-video and reference-to-video generation. It can animate a single still image and create automatic scene transitions, giving developers two input paths for different product workflows. Atlas Cloud provides dedicated API endpoints with published parameter schemas at the verified standard price of $0.04 per second. Start building today.

View Family

Gemini Omni 1.1 Flash

Gemini Omni 1.1 Flash is Google DeepMind’s natively multimodal video family for flexible production workflows. Edit existing footage through text instructions, maintain visual consistency with up to 10 reference images and 3 video clips, or extend a shot in chainable segments to 40 seconds. Access every workflow on Atlas Cloud with one OpenAI-compatible API key, pay-as-you-go pricing, and Day-0 availability. Start building today.

View Family

Nano Banana 2.1

Nano Banana 2.1 is Google's image model family built on Gemini 3.6 Flash. It delivers Pro-level generation and editing at Flash-level speed, supports natural-language edits and merges across 1 to 14 input images, and maintains consistency for up to 4 characters and 10 objects. Through Atlas Cloud, developers can access the family with one OpenAI-compatible key, reliable uptime, and Day-0 availability. Start building today.

View Family

Nano Banana 2

Build with the Nano Banana 2 API, powered by Google's Gemini 3.1 Flash Image model. It generates native 4K visuals up to 4096x2304 with accurate text rendering and consistent characters across up to 14 reference images for both generation and editing. On Atlas Cloud you reach it through one unified API alongside 300+ models, with pricing from $0.04 per image, 99.99% uptime, and free credits to start.

View Family

One API for All Media AI.

Explore all models