Small, fast and cheap
Rosenblatt Perceptron
The simplest learning unit. Built for high-volume, low-latency work and on-device use.
Perceptron is our smallest model. It answers quickly and costs little, so you can run it on every request, every keystroke or every row of a dataset.
It's small enough to run on-device, which keeps latency low and data local.
- Speed
- Fastest
- Reasoning
- Direct answers
- Access
- API
- API name
- rosenblatt-perceptron
Good for
- Code completion
- Classification and extraction
- Quick checks at scale
- On-device features
Other models
Balanced, the everyday workhorse
Layer
A multi-layer model for most coding, writing and analysis work.
Learn more Flagship, most capableDeep
Our top model for hard reasoning and long, complex tasks.
Learn more Research tier, limited accessMark
A frontier model for specialized or sensitive work, with tighter access controls.
Learn more