Editor's note
What is live, how to access it, and which preset lanes we recommend if you want to try the new model without rebuilding your stack.
New model launch: available on day zero with routing presets
Model launches move fast, but operational adoption usually moves slowly. Our goal is to close that gap by shipping a usable route on day zero, not just a logo in a catalog.
Try the new model without reworking your integration.
Launch from UI presets or send traffic through a route that can be rolled back instantly if quality or price does not hold.
Launch Team
What is live, how to access it, and which preset lanes we recommend if you want to try the new model without rebuilding your stack.
What shipped
- Direct API access through the unified gateway
- Preset lanes for quality-first, speed-first, and cost-aware evaluation
- Usage logging and billing attached to the same control surface
Why this matters
A model is only truly available when teams can test it under real traffic conditions, compare it to existing defaults, and revert without drama.
Platform News
Try the new model without reworking your integration.
Launch from UI presets or send traffic through a route that can be rolled back instantly if quality or price does not hold.
Related reads
Back to blogKeep the same thread going with adjacent pricing notes, launch posts, and hands-on walkthroughs.
Product5 min
Spring 2026 product update log
A readable changelog that groups improvements by the workflows people actually care about: routing, workspace, and creator output quality.
Product7 min
Case study: from one idea to a full campaign system
How a small team used shared prompts, reusable presets, and multi-model routing to go from concept to launch assets in two days.
Developers7 min
How to call 125+ AI models with one API key
A practical architecture for teams that want one billing layer, one request format, and room to swap providers without touching product code.