SurgePix
The "agent-callable" line is doing all the heavy lifting here, and good for them. We're watching every SaaS tool on earth race to bolt on an API for AI agents, and SurgePix seems to have started there instead of tacking it on later. That's a meaningful difference in posture. If you're building for the agent economy, your product's accessibility to another AI system matters as much as its interface for humans.
Layer-level editing is the other hook worth paying attention to. Most AI image generators hand you a finished picture and dare you to argue with it. Per-layer control means compositing, selective edits, maybe even a pipeline where an agent tweaks just the background without nuking the subject. I'd want to know how deep that layer stack goes and whether it exports in formats that fit into existing design workflows.
I'm not going to pretend I have a read on traction or team, because there's nothing in front of me beyond the tagline. But I'd love to see a comparison against the established players, because the market has gotten very noisy very fast. Who's the actual customer here? If it's developers wiring visuals into their products, SurgePix has a shot. If it's hoping to win designers over with another UI, that's a tougher sell. The agent-callable framing suggests they've already picked their lane, and I respect that clarity.