Happy 4th of July from the Nebula team!
While everyone's launching rockets, we shipped:
> 162 PRs
> 12 features
> 153,255 lines of code
...in the past week.
Just a couple humans and a ton of AI Agents.
Upwards and onwards!
Fable 5 is back. Sonnet 5 dropped.
So we added a Claude Code integration into Nebula.
Your agents can now ship code and build apps using your existing Claude plan. No extra token spend.
Happy tokenmaxxing,
The Nebula Team.
Nebula can orchestrate claude code for you now, including using your existing plan.
It works for all code related tasks and runs on your nebula device.
It’s mindblowing to me that LLMs are the fastest growing consumer product of all time
…and the “multiplayer” experience (across every provider!) is basically just being able to share a chat history
Someone is going to crack this and it’s going to unlock a LOT more usage
We pay our newest employee $4.50/mo.
Yes, a month.
He's going through our GitHub issues, support emails, and tickets - then ships improvements.
This is our new product lead, and he's an AI Agent:
Our team ships to prod from their phones.
This is the biggest unlock of using AI Agents.
Since you can talk to them from anywhere, your laptop becomes optional.
From our phones, we've:
> Built and QA tested a feature
> Created and shipped videos
> Ran paid ad campaigns
Who else is running their business from their phone?
Try out this workflow ASAP.
DM us and we will connect you to our video editor to help get you started.
P.S. this video was created by the same person who made our funny stop motion ads.
Figured out an insane workflow for creating Seedance 2.0 videos
> Use @NebulaAI and create an agent that helps you write prompts for image and video.
> Create reference images on @higgsfield, edit with GPT 2 image.
> Generate image-2-video on Nebula.
The results are crazy.
Everyone's waiting for Fable 5 to come back.
But your agents don't need better models. They need cheaper ones.
Fable 5 costs $10/$50 per million tokens.
One agent running 50 heavy autonomous calls a day costs ~$13.
Do that with 10 agents and you're at $130/day.
Nebula's new model costs $0.75/$3 per million tokens. Same agents cost ~$9/day.
Of course, Fable is amazing for certain tasks.
But LLMs are already capable enough. The bottleneck is no longer compute. It's economics.
The real question is - are you building the right agents to take advantage of them?