d1-3B

Liquid AI · released Oct 7, 2026

Unknowntextvision

Openness taxonomy

access, licence, and disclosure — never a single open/closed binary

Weights availability: Open-weight and available on Hugging Face
Licence: unknown
Licence category: unknown
Commercial use allowed: unknown
Derivatives allowed: unknown
Redistribution allowed: unknown
Field-of-use restrictions: unknown
Training code availability: unknown
Inference code availability: Models ship their own code; loading instructions use trust_remote_code=True
Architecture disclosure: Built on Liquid Foundation Models; described as a single-forward-pass decision model
Training data disclosure: unknown

openness confidence

Release states both decision models are open-weight
Release states models are available on Hugging Face

Deployability

raw values as extracted — free text, never inferred from branding

Self-hosting: unknown
On-premise: unknown
Edge deployability: Evaluated on GeForce RTX 4090, Jetson AGX Thor, Jetson AGX Orin 64 GB, and Jetson Orin Nano
API availability: unknown
Regions available: unknown

Benchmarks

as reported — vendor-supplied unless otherwise noted

Decision Index v0.3 mean score across seven public datasets 82.9Edge inference, single question on measured devices under 50 ms on every measured deviceGPU inference, one question under 10 msGPU inference, 384px image under 18 ms on both platforms

National-security pathway & evidence

from the assessed source item, when one is linked

Impact: mediumConfidence: mediumNat-sec: noneprimary

Evidence limitations: Performance and benchmark claims are reported by the release source; no independent evaluation is provided. The post says it reports no vision or audio benchmarks and gives no speed numbers for d1-omni-600M.

Multimodal open d1 decision models for the edge

Linked research papers

from paper_model_links

No linked research papers yet.