The Creator Workspace
Image, video and audio in one composer. Chain any result into the next step, redo past runs in one click, never re-upload a file.
Built for Volume, Not One-Off Generations
The classic tool pages are built around a single operation: pick a model, fill a form, get one result. The Creator is built around the session instead. A floating rail switches between image, video, and audio and every action each one supports; a single composer at the bottom handles prompts, files, and parameters for whichever operation is selected; and every result - past or present - lands in one canvas you can keep working from. Generate a batch of product images, chain the best one into image-to-video, add an AI voiceover, and pull a matching template in along the way, without a single download-and-reupload round trip.
Every classic single-operation tool, rebuilt around a session you can keep building on.
The Workspace
1.
One chat-style input bar handles every operation - image generate, edit, upscale, inpaint, outpaint; video text-to-video, image-to-video, avatar, motion control; speech and music generation. Switch mode and action from the floating rail without ever losing your place.
2.
Turn any result into the next input with a single click. Generate an image, chain it straight into image-to-video, then chain that clip into an upscale or a subtitle pass - all without downloading and re-uploading a single file.
3.
Every generation in the current session lands in one scrollable grid. Multi-select tiles to merge clips, batch-remove, or re-select the exact source assets a merge was built from.
4.
Every past generation is one click away. Pin a result from history straight back into your session, or hit Redo to reload its exact model, prompt, parameters, and input files - ready to run again or tweak.
Browse a library of preset prompts and parameters across image, video, and audio. Selecting one instantly configures the composer - model, prompt, and settings included - so you start from a proven result, not a blank box.
Generate up to 4 image variations from a single prompt in one request, so you can pick the best result without repeating the same submission four times.
Enhance a rough prompt into a detailed one with a single click, or reverse-engineer the prompt behind any image you've generated or uploaded - both built directly into the composer.
The Creator routes each operation to the model you pick - no separate pages, no separate accounts to manage.
Flux 2, Flux 3, Nano Banana 2, GPT Image 2.5, Krea 2, and Seedream 5 - generate, edit, upscale, inpaint, outpaint, and remove backgrounds without switching providers.
Veo 3.1, Sora 2, Kling 3.0, Seedance, and Happy Horse - text-to-video, image-to-video, avatar generation, and motion control in the same composer.
ElevenLabs, OpenAI, Gemini TTS, Qwen 3 TTS, and MiniMax Music - speech and music generation alongside your visual assets, in the same session.
ElevenLabs
OpenAI TTS
Gemini TTS
MiniMax MusicWhenever one generation is never enough, the Creator is where the work happens.
A studio needs 40 ad variations for a client review by end of day. They generate a batch of 4 images per concept across a handful of prompts, chain the strongest concepts into image-to-video, and keep every version in the session board for side-by-side comparison - no juggling separate tool pages per format.
A social team turns one product photo into a week of content: an image edit for a clean background, a chained image-to-video for a reel, and an AI voiceover generated alongside it - all three built in sequence from the same composer, without leaving the page.
A marketer redoes last quarter's best-performing creative with a new product angle. Redo reloads the exact model, prompt, and parameters from history, they swap in the new product photo, and a fresh batch is ready in minutes instead of rebuilding the brief from scratch.
A solo creator sketches a concept with a quick generation, uses "Get Prompt" to reverse-engineer a reference image they liked, tweaks it with Enhance, and chains the result through three operations before landing on a final asset - one continuous workspace instead of five separate tabs.
One workspace, every operation - here's what's under the hood.
The classic pages (image/generate, video/edit, and so on) are each a single form for a single operation. The Creator is one continuous workspace: a floating rail switches between every operation across image, video, and audio, one composer at the bottom handles the current one, and every result - current session or past history - lives in the same canvas you can keep building from.
Stop rebuilding the same generation from scratch. Chain, redo, and template your way to volume.
The Creator includes
