Lights, Camera, Traction
Introducing Yarn – the video product for agentic GTM teams.
The growth teams at Clay and Faire have made 100s of videos with Yarn already.
You provide the spark, agents do the rest. It’s Waymo for video production.
Our launch video was made entirely in @yarn_so
Lights, Camera, Traction
Introducing Yarn – the video product for agentic GTM teams.
The growth teams at Clay and Faire have made 100s of videos with Yarn already.
You provide the spark, agents do the rest. It’s Waymo for video production.
Our launch video was made entirely in @yarn_so
Unfortunately the reasoning model prices vary and we pass through savings immediately (and then there's input vs. output, cache-reads vs. not). There are also different secondary agents within Yarn that have different model-pricing.
So I'm not sure that would be much clearer.
We implement clear buckets within the app to prevent accidental over-spend, and many users have been able to successfully complete fairly long videos in their free allowance!
I do think it's a tricky problem with reasoning agents today, and hear you that we can do a better job at trying to show what credit usage can yield.
As more teams use Yarn, hopefully we can have a better picture on that!
Lights, Camera, Traction
Introducing Yarn – the video product for agentic GTM teams.
The growth teams at Clay and Faire have made 100s of videos with Yarn already.
You provide the spark, agents do the rest. It’s Waymo for video production.
Our launch video was made entirely in @yarn_so
Yep agree on both!
We're really excited by contextual/concierge content i.e. show me relevant content just for me - more than "mail merge" style personalization.
I do think you can go very very far in this direction without trading off any authenticity!
But definitely requires lots of taste and more systems-thinking around content :)
Hey Krisof!
Our customers don’t use a lot of image or video gen models today – which as you know can be cleanly priced per image/second of video!
Yarn’s agent uses frontier reasoning models primarily, which are priced per tokens.
So it really depends on the complexity of the content, your quality bar, and how much automation you want to do.
If you use max thinking and add lots of agent comments, the cost will be higher than if you do lots of the work manually/by hand.
As an example, that launch video cost approximately $300 in tokens. I could’ve made it more manually and with fewer-comments/lower quality bar for under $25.
As our customers make more content, our harness becomes more token efficient (i.e. has to do less thinking per new piece of content).
@dontbanjake
@gleblapkovsky Yep that production pipeline is really amazing. Hopefully creators will spend more time on storytelling and raising the ceiling with e.g. personalization!
Yep fair we could be clearer. This is our current guidance!
2,500 credits - Explore the product
10,000 credits - Agency-quality shorts
20,000 credits - Ship videos every week
40,000 credits - Scale content across GTM
80,000 credits - Automate most of your content
Most teams switch from using agencies which might be $4k+/video at the low end.