vivieches/open-studio

Open Studio is a local first AI content workspace for video creators. The core idea is to collapse a typical creator's scattered workflow, scripts in one tool, thumbnails in another, captions in a spreadsheet, title ideas somewhere else, into a single application that treats the title, thumbnail, script, and caption as one connected delivery package instead of separate random outputs. You start with a brief or topic. Open Studio uses AI providers to generate a script tailored to your audience and tone, produce a batch of real thumbnail images, create ten SEO and click rate optimized title candidates ranked down to a top three, and write a caption with hashtags, keywords, links, and social blocks. A pipeline view runs the whole sequence in order: brief, script, titles, thumbnail, captions, critique, export. Generated assets and exports are stored locally rather than in the cloud. The app supports multiple AI providers. You can bring your own API keys for text and image generation, use locally installed AI command line tools such as Claude Code, Codex CLI, or Gemini CLI that it detects on your machine, or connect a local language model. There is also a free option for basic image testing that needs no API key. Three interface languages are supported: Portuguese Brazil, Spanish Spain, and English US, and the app writes image prompts in English while keeping any visible text in the image in the chosen language. Open Studio is built for creators who want a real production tool rather than a simple mockup, and it includes a standalone local daemon for run tracking, health checks, and logs. The tech stack is TypeScript and Next.js, and it requires Node.js 20 or newer to run.