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, without sampling. 1/6
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, without sampling. 1/6
Today we're launching Mirelo SFX 1.6 - our first model that lets you edit sound, not just generate it. It's a meaningful first step toward full iterative sound editing.
We're excited to share with you:
Long Generations: generate up to 60 seconds of audio from video.
Extender: extend any sound with seamless, natural tails or intros.
Ambiancer: generate soundscapes that naturally end where they begin.
Inpainter: clean a sound clip by replacing unwanted sounds.
Read more in our blogpost in comments. SOUND ON.
Excited to share that @lmnrai has raised $3M to build open-source observability for long-running AI agents.
Laminar is how companies like @browser_use, @OpenHandsDev, and Rye see what their agents are doing, understand why they fail, and spot patterns across millions of runs.
@yrechtman Try to grip lower in your hand. You will feel weaker at first but it’s worth it. See here the legend Rippetoe explaining it (with his amazing accent): https://t.co/YKU3p4UaRe
@laserlikemike Check out @AnnieDuke ‘s new book “Quit” for a more differentiated take: you need to quit things in order to be able to try out many things.