Veo 3.1 Quality
High-quality 8-second text-to-video and first-frame generation at 720P, 1080P or 4K. Image reference mode is not supported.
Glossary & Docs
Dream AI glossary and usage docs for AI image generation, AI video generation, multi-model hubs, Seed locking, image-to-image, image-to-video workflow, and delivery-ready parameters.
High-quality 8-second text-to-video and first-frame generation at 720P, 1080P or 4K. Image reference mode is not supported.
Lower-cost 8-second text-to-video at 720P, 1080P or 4K. No image input.
APIMart extended Gemini Omni endpoint. 4/6/8/10-second generation at 360P, 720P, 1080P or 4K, with first-frame control and exactly 1 or 3 reference images.
4–15 second generation at 768P or 2K with an audio track. Dream AI supports text, first-frame control and up to 5 image references.
1–15 second text-to-video, first-frame and 1–7 image reference fusion at 360P, 540P, 720P or 1080P. Dream AI uses silent output and supports seed and watermark.
5 or 10 second text-to-video and first-frame generation. 720P uses Standard mode; 1080P uses Pro mode. Dream AI uses silent output.
3–15 seconds at 720P / 1080P, with first-frame control and up to 9 reference images.
Standard text-to-image; Pro reference image editing.
A newer GPT Image generation and editing model listed in current provider catalogs. Exact endpoints and capabilities can vary by provider.
A recent Wan video-generation model listed by provider catalogs for prompt-led and reference-based video workflows.
A newer Nano Banana image model listing for generation and editing; provider implementations may differ.
A professional-tier Seedream image model listing for detailed generation and editing workflows.
A newer Kling video model family entry listed by provider catalogs. Model modes and controls depend on the service.
Fast, Quality and Lite APIMart endpoints; fixed 8-second clips at 720P, 1080P or 4K. Fast supports first-frame and up to 3 reference images; Quality supports first-frame; Lite is text-only.
A recent LTX video model listing for video-generation workflows. Exact supported inputs vary by provider.
A Vidu video-generation model family entry listed in current provider catalogs.
APIMart extended video endpoint using the Grok model, with 6–15 second text-to-video, image-to-video and up to 7 image references, at 480P or 720P.
FLUX 3 is Black Forest Labs' multimodal foundation model, jointly trained across images, video, and audio to support unified visual creation and world understanding. It is currently in Early Access and is not yet available for generation in Dream AI.
Nano Banana 2 Lite is Google's fastest and most cost-efficient Gemini Image model, designed for rapid ideation and high-throughput image generation and editing.
Seedance 2.5 is available through APIMart in Dream AI for text-to-video, first-frame image-to-video and reference-image generation, with clips up to 30 seconds and native audio.
Seedance 2.0 Mini is presented in public product listings as a lighter, faster, lower-cost Seedance 2.0 variant for drafts and repeated experimentation. Exact availability and specifications remain unconfirmed by official ByteDance documentation.
Grok 4.2 is positioned in Dream AI for stronger semantic understanding, character movement, and composition.
Ideogram 4.0 is listed in Dream AI as a coming-soon model for posters, logos, layouts, and typography-oriented image generation.
Qwen Image 2.0 is a general-purpose Dream AI model with strong Chinese prompt understanding and text-rich visual capabilities.
GPT Image 2 is positioned for stronger prompt understanding and realistic rendering in Dream AI.
HappyHorse 1.0 is a Dream AI video model for text/image-to-video generation and camera motion workflows.
Wan 2.7 is a high-realism Dream AI image model with Chinese aesthetics and detailed realistic rendering.
Grok Imagine is a creative Dream AI model for expressive composition, artistic style, and high-energy visuals.
Nano Banana 2 is the default fast-draft image model in Dream AI for prompt-to-image and reference-based iteration.
SkyReels V4 Fast prioritizes video generation speed for text-to-video and image-to-video workflows.
SkyReels V4 Standard focuses on cinematic camera motion and visual texture for text-to-video and image-to-video generation.
Seedream 5.0 is Dream AI's cinematic image model for physical lighting, material realism, and polished visual texture.
Seedance 2.0 Standard balances visual quality, motion range, and multimodal video generation in Dream AI.
Seedance 2.0 Fast is optimized for faster turnaround in Dream AI video generation.
Seedance 2.0 Face is optimized for portraits, facial detail, and character motion in AI video.
Seedance 2.0 Fast Face combines portrait optimization with faster generation speed.
GPT Image 1.5 is the classic stable GPT Image option for general image generation workflows.
Seedream 4.5 is a versatile Dream AI option for illustration, Chinese-style visuals, and general realistic generation.
Z Image Turbo is a speed-oriented Dream AI model for fast single-image generation.
Nano Banana Pro emphasizes finer detail, richer color, and a more polished rendering style than the fast draft variant.
Flux.1 Kontext Pro is optimized in Dream AI for long prompts and complex multi-subject layouts.
Flux 1.1 Pro is a premium realism model in Dream AI for commercial-quality faces, detail, and lighting.
We are creating exciting content, please stay tuned!