Create Studio
Create images and videos directly. Write a prompt, add a reference, choose a model, and work with each result in a familiar project feed.
AgentDocks documentation
Learn AgentDocks from the beginning.
AgentDocks is an AI visual production platform for creating images and videos. Work directly in Create Studio, or build a connected process in the Node Editor when you need more control.
The platform
From a first idea to finished visual work.
A single generation is often only the first step. AgentDocks keeps the inputs, creative decisions, tools, and results together so you can continue the project instead of starting again each time.
01
Start with an idea or reference
Describe what you want to make, upload something to work from, or begin with a prepared template.
02
Build the work in one project
Your prompts, inputs, settings, and results stay together while you create and make changes.
03
Keep the process when it works
Move into a workflow when you want to see the steps, change the logic, or run the process again.
Two ways to work
Choose the interface that fits the job.
Create Studio and the Node Editor use the same production environment. Neither one is a beginner or professional edition of the other.
Node Editor
Build the process as connected steps. Use it when you want to see how information moves, branch a direction, or run the same setup with new inputs.
Templates help you start ahead.
A template gives you a prepared starting point. Add your own inputs, run it, and change the underlying process when you need something more specific.
Keep learning
Find the next useful answer.
Follow the guides in order, or open the exact page that matches what you see on screen.
Start here
Create Studio
Templates
Node Editor
- Canvas basicsAdd nodes, wire ports, read wire colours, tools, shortcuts, undo and autosave.
- Run workflowsRun one node, selected nodes, or the complete workflow.
- Batch Run and triggersRun a branch once per picture, and start workflows from a schedule or a Telegram chat.
- Telegram remote controlCreate a bot, connect and pair it, send from your phone, fire with a word, get results back.
Nodes
- Image GenerateGenerate images from prompts, references, and connected inputs.
- Video GenerateGenerate clips from prompts, references, a start frame, or an existing video.
- Motion ControlTransfer a reference video's movement onto your character image.
- Text GenerateChat with a language model on the canvas and hand its text to other nodes.
- Model GeneratorA photoreal human model from options or references, the same person from every angle.
- Camera AngleRe-photograph one image from a new camera position; the scene stays frozen.
- Image to PromptAn art director that writes generation prompts from a reference image.
- Character SheetOne character into a labeled sheet of views, expressions and poses.
- Image toolsUpscale, Remove Background, Crop & Resize, Describer, Prompt Enhance, Color, Palette, Painter, Design Studio.
- Video toolsTrim, Extract Frame, Video Describer, Alpha Matte, Video Color Match, Video Upscale.
- Color CorrectionLevels, Curves, Color Balance, LUT, Filters and the rest, free and on-device; Color Match by model.
- Inputs and previewsImage, Video, Text and File inputs, the previews, and Note and Frame.
Account
Troubleshooting
- Run did not startCheck validation, model pricing, credits, and active runs.
- Reference ignoredModel support, where the picture sits, the limit, how the prompt names it.
- Run stuck or slowQueued and running, how long is normal, stopping and refunds, reloads.
- Upload refusedFormats, size limits, storage, the model's own limits, template fields.
Start with something real