Hey Sumo-Lings! 👋
I'm Amogh, and I am extremely excited to launch "Friday Code" on AppSumo!
We built Friday Code because there isn't any desktop agent which is specifically built for running your entire product.
Why is it "Friday"? It's inspired by Tony Stark's assistant "Friday", we all needed one of that.
With Friday, you can build full stack apps, mobile apps, websites, blogs, email campaigns, and automate with MCP integrations such as Supabase, Vercel Figma, Slack.
Bring your existing projects from Lovable, Replit, Github, Claude Code and start working on them in Friday.
Here are some of the cool features:
- Share your screen and specific apps, so that Friday can takes action based on what it sees. You can clone websites or let Friday improve your existing product.
- Build real apps, complex software with backend. You can setup realtime database, create edge functions with Supabase, setup Github workflows, deploy to Vercel.
- Bring in your ideas, images, figma, existing project and get started.
- Use Automations to let Friday run on a schedule, to handle support requests, write blog posts, send out email campaigns, fix bugs and much more!
Use purpose-built Friday Agents for your tasks:
- Friday Go: Best for fast, general purpose tasks, simple apps, websites and fixes.
- Friday Coder: Ideal for complex projects or full stack apps.
- Friday Writer: Ideal for writing heave tasks
- Enable "Long Running Agents" which can autonomously run for several minutes or even hours, depending on the tasks. Agents can search the web, communicate with your connected apps, write and publish code and do much more!
How does the Credit system work?
- All the tasks use credits.
- Credits are charged based on the amount of tokens each query uses and the action it takes. Since the agent automatically takes actions and completes a task, it has to make multiple queries, ranging from dozens of queries to hundreds of queries per task.
- Each Agent has a credit rate set and it is visible in the UI. We recommend using the cheapest possible agent for simpler tasks. All agents use state of the art AI models.
- We recommend checking the credits used for your tasks, so that you can make informed decisions on how long you want the agent to run and which agents are ideal for your tasks.
- You can also select 300+ models via OpenRouter with your own API key, the models include all the Claude, OpenAI, Gemini, Grok, DeepSeek, Kimi, Qwen and other models. This allows you to pick the best models for your use.
We ship new features almost every day!
Please join our Slack, and say hello, we'd love to learn more about what you're building!
Describe your idea in plain language and watch Friday Code generate a working, full-stack application
