Studio Workspace Overview
Redbit AI is a creative production workspace for turning prompts, source media, and reusable assets into image, video, speech, audio, and project outputs. The core workspace is built around independent task cards, an Asset Dock, a Creative Workshop, a guarded Agent layer, and explicit model/provider configuration.Who Should Read This
Before You Read
If the product terms are unfamiliar, skim Concepts and Glossary first. Redbit coordinates workflows around configured providers; it does not guarantee provider quota, latency, uptime, policy, or output quality.Recommended Reading Path
- Studio Workspace Overview
- Evaluation Guide
- Quickstart
- First Project Walkthrough
- Decision Matrix
- Core Workflows
- Feature Modules
- Models and Provider Configuration
- Security and Credentials
- Agent Workflows
- Seedance Video Workflows
- Troubleshooting Playbook
Product Positioning
Redbit is not only a chat interface and not only a model API wrapper. It is a workflow layer that keeps creative work visible and reusable:- prompts and generated media live in cards instead of disappearing into one transcript;
- assets can move between cards, Asset Dock, SmartPicker, and Workshop projects;
- model routing is configured through Settings and the model registry instead of hard-coded per button;
- the Agent can inspect workspace state and call bounded tools for cards, assets, Workshop projects, settings, search, and local integrations.
Main Areas
1
Canvas
The
/dashboard workspace shows cards for image, video, speech, describe, and audio tasks. A card keeps its own prompt, model, variant, ratio, references, generation status, and results.2
Asset Dock
Asset Dock stores reusable image, video, audio, and text assets. Items can be pinned, searched, compared, and selected through SmartPicker as references for later work.
3
Creative Workshop
/workshop manages project-level creative production: script, storyboard scenes, image generation, video generation, voiceover, music, rhythm video, consistency references, and export packages.4
Settings and model registry
Settings control model defaults, provider credentials, relay behavior, custom model endpoints, Agent runtime mode, and capability testing. The catalog defines supported model families, variants, ratios, and input requirements.
5
Agent panel
The Agent reads a structured workspace snapshot and calls registered tools. Destructive operations and external effects are bounded by tool metadata, result normalization, and explicit target checks.
