I like that we go back to idea for cost-efficient specialized models
I am tired to hear 'but my claude can already to that', sure it can do everything. But you don't know for what price because you don't pay for it. Your employer pays!
And token cost if what matter to business
Introducing flow-1, our new model trained with RL to find errors in agent traces.
It matches GPT-6-sol in trace intelligence while being 23x cheaper. It also costs 25% less to run than GPT-6-luna.
flow-1 finally makes it possible to monitor and understand every agent run,
00:00
