Added Bytedance Upscaler for clip upscaling with selectable target resolution (1080p, 2K, 4K).
Adds FireRed Image Edit, a multi-image editing model re-trained from Qwen Image Edit with support for English and Chinese instructions, virtual try-on, and style transfer.
Edit videos with Kling O3 Standard and Pro — reference images, character/object elements, and keep-audio support.
A new dedicated tool for AI image generation—model selection, reference images, Cinema camera controls, AI prompt assistance, and one-click video creation.
Adds Pruna P-Image as a text-to-image model—sub 1 second generation with state-of-the-art quality, exact prompt adherence, and exceptional text rendering.
Adds Kling v3 Standard and Kling v3 Pro with text-to-video and image-to-video support, automatic endpoint routing, optional native audio, and 3–15 second durations.
A dedicated Video Generator tool with an intuitive workflow for creating videos from prompts, images, AI-generated frames, and reference media.
Adds ImagineArt 1.5 Pro Preview for text-to-image generation—ultra-high-fidelity 4K visuals with lifelike realism and professional output.
Define your brand's visual identity and ad rules in one place. Colors, fonts, logos, voice, layout, and ad-creation rules—so the AI stays on-brand when generating videos, rich media, and static ads.
Adds Kling O3 I2V and Kling O3 Reference for image-to-video with optional start/end frames, reference elements, and synchronized audio. 3–15 second clips.
Adds xAI's Grok Imagine Image for text-to-image generation and image editing—highly aesthetic output, style transfer, and precise edits from natural language prompts.
Adds xAI's Grok Imagine for video generation and editing—text-to-video, image-to-video with optional audio, and video editing for restyling, object control, and scene refinement.
Generate responsive HTML5 ads from a business website. Pick a business, configure ad objective, style, language, and IAB sizes—the AI creates rich media ads optimized for display.
A new prompt-driven playground for creating longform videos. Enter a topic, add reference images or videos, choose language and audio style—then let the AI generate 15-, 30-, or 60-second videos.
Adds ByteDance's Seedance 1.5 Pro for joint audio-visual video generation—sound and vision in one take, with cinematic camera control and multilingual support.
Adds Black Forest Labs' FLUX.2 [max] as an image model for generation and editing—highest quality in the FLUX.2 family with strong character consistency and text rendering.
Adds Bytedance seedream-4.5 as an image model for generating static ads—with superior aesthetics, better text rendering, and strong reference-image consistency.
Adds Kling v2.6 Pro for image-to-video with optional native audio, GPT-5.2 for the AI editor agent, and improves video editor and preview handling.
Upload images and videos as 'Video Ingredients' to guide AI video generation. Available in all playgrounds and the create-video dialog—drag-and-drop your attachments and they flow through to generation and the editor.
We've integrated Kling O1 to enable character-consistent video generation with reference images and elements.
We've added a new feature to automatically generate masks for your images using SAM 3.
We've added a new Standard V2 Preview agent mode to the playground and video AI.
Nano Banana Pro is now available as an image editing and generation model for your video workflows.
Gemini 3 is now available as an LLM option for your video workflows.
We've added bumper video durations to the video AI workflow.
Seedance Fast is now available as a video generation model for video AI workflows.
We've integrated the ElevenLabs v3 model to improve the quality of the audio generated by the video AI.
We've added UGC product ads to the platform, allowing you to generate product ads using AI.
We've integrated the Reve create/edit model to increase our image generation toolset.
We've integrated the Sora 2 model to increase our video generation toolset.
We've added AI influencer generation to the platform, allowing you to generate influencers using AI.