Fine-grained workspace controls
Adjust subtitles, shots, audio, and generation settings precisely.
The open-source AI workspace for creators
From scripts to video, images, voice, avatars, translation, and editing, Agents move the whole creative process forward in one workspace.
Video translationSmart editingThumbnail generation
Image generationAI voiceoverVideo download
Every creative tool, in one workspace
Each workspace shares project context, version history, and output files.
Conversation and workspace, moving together
Describe tasks naturally, then step into visual tools whenever you need precise control.
Flexible conversational edits
Tell the Agent what to change and refine the result in natural language.
Synchronized state
Conversation and workspace share the current task state, so nothing needs repeating.
Independent versions
Each revision creates a separate version without overwriting earlier results or settings.
Start with a template. Build on it with Skills.
Find a starting point for your next idea, then make the methods you use most reusable with your Agent.
Creation templates
Skip the blank prompt
Explore image and video templates created by our team and curated from popular third-party sources, with example results, prompts, and settings. Pick a starting point and make it your own.
Choose a template Adjust settings Create
Agent Skills
Give your Agent a way to work
Reuse workflows for subtitles, dubbing, covers, and more. Write your own SKILL.md to connect the tools and services you already use.
Install the Skill and configure any services it needs before use.
skills/SKILL.md
krillinai-subtitleSubtitles & translation
krillinai-ttsVoice & dubbing
krillinai-coverCover creation
The right model for every creative task
Codex powers the Agent. Language models come from the Codex catalog or a compatible provider; creative services are configured separately. These are examples, subject to your account, service setup, and platform.
Language models
- GPT
- DeepSeek
- Qwen
- Kimi
- GLM
- Grok
- Doubao
- ERNIE
- Hunyuan
- MiniMax
Image
- GPT Image
- Seedream 4.0
- Kling v2.1
- Nano Banana
Video
- Seedance 2.5
- Kling v2.1 Master
- Veo 3.1
Voice and transcription
- Whisper
- OpenAI TTS
- MiniMax Voice
- Edge TTS
- Aliyun Speech
Local transcription also supports faster-whisper, WhisperKit, and whisper.cpp. Video translation supports Volcengine ASR / TTS.
An open ecosystem, built with creators
We create characters with designers and share templates and workflows, so creators can reuse, improve, and contribute new capabilities.
Built by theopen source community
Every contribution, discussion, and shared workflow makes the creative toolkit better.
- GitHub Stars
- 12.0K
- Contributors
- 21
- Countries represented
- 98
Loading contributor avatars.
Your creative data stays local
OpenCreator runs locally, builds on Codex Agent capabilities, and keeps growing through Skills and MCP.
Open source · Apache-2.0Local first
Projects, attachments, and runtime logs stay on your device by default.
Built on Codex
Reuse models, tool calls, conversations, and approval flows directly.
Designed to expand
Connect new creative services through Skills and MCP.
One Web and Desktop experience
The same frontend and Runtime keep behavior consistent across platforms.
Frequently asked questions
Quick answers about setup, data, and model services.
What is OpenCreator
OpenCreator is an open-source AI workspace for creators that brings Agents, video, images, voice, translation, and smart editing into one creative workflow.
Which operating systems are supported
Installers are currently available for macOS on Apple silicon, macOS on Intel, and 64-bit Windows 10 or 11. A Linux desktop build is not yet available.
Does my creative data leave my device
Projects, attachments, and runtime logs stay local by default. When you use an online model, the content required for that request is sent to your configured model provider.
What do I need to get started
After installing the desktop app, sign in to Codex or configure the model service used by the Agent, then add any required API Key, Base URL, and model under AI Services. Some services, including Edge TTS, do not require an API Key.
Can I use my own model provider
Yes. Text, image, video, voice, and transcription services can be configured independently, including services compatible with the OpenAI API.
Is OpenCreator open source
Yes. OpenCreator is licensed under the Apache License 2.0, and its source code is available on GitHub. Third-party models and external services may charge according to their own pricing.