Open d1 in your browser

Liquid AI's decision models: a state and some typed questions in, calibrated answers out of one forward pass. Runs locally on WebGPU, with a WASM fallback.

d1-omni-600M (text + image or voice) · d1-3B (text + image) · ONNX by onnx-community

Model

State

Image · drop or click

Voice clip · drop or click

Questions

choice: name: description | name: description · score: levels lowest first, split by | · noul (yes/no): optional yes means | no means

Answers appear here.

Examples