Decision models are the edge play nobody markets like chatbots. liquid ai's d1-3b and d1-omni-600m,…
decision models are the edge play nobody markets like chatbots. liquid ai's d1-3b and d1-omni-600m, out this week, run fast classifications on a phone instead of calling a data center.
the next model war is in the output nobody sees as text.
Context
Liquid AI's blog post of October 7, 2026 announces d1-3B and d1-omni-600M, two open-weight models in its d1 decision model family. Liquid's documentation describes decision models as returning calibrated probabilities across a fixed set of outcomes in a single call with zero generated tokens, for classification, routing and scoring.
The post says d1-3B scores 48.57 on the Decision Index v0.2.1 public split, ahead of every model under 10B and on par with a decision model 12 times its size. It reports latency of 8 ms on an NVIDIA GeForce RTX 4090, 16 ms on a Jetson AGX Thor and 26 ms on a Jetson AGX Orin, and says it measured on an Apple M5 Pro as well. The Hugging Face card calls d1-omni-600M experimental.
The release date and the two model names match the Liquid post, which is dated October 7, 2026. The Decision Index is Liquid's own benchmark, so the 48.57 score is a vendor figure, not an independent result.
The note says the models classify 'on a phone'. The pages read report latency on an RTX 4090, Jetson boards and an Apple M5 Pro laptop chip, not on a phone, so phone performance is unsupported here, not refuted.
'Decision models are the edge play nobody markets like chatbots' and 'the next model war is in the output nobody sees as text' are the author's lines. The documentation does support the narrower point that these models return probabilities and generate no text tokens.
Related work
- Open d1: Edge decision models for text, vision, and audio (Liquid AI blog, October 7, 2026) ↗Source for the release, the Decision Index score and the device latencies.
- Decision Models (Liquid Docs) ↗Liquid's description of decision models and their single-call, zero-token output.
- LiquidAI/d1-3B (Hugging Face) ↗Model card for d1-3B.
Watch next
- Look for independent evaluations of d1-3B outside the Decision Index. Check whether any phone latency has been published.
Sources
Provenance
The note above is reproduced unedited from the original post, first published on Threads on 9 October 2026 at 09:37 IST. Sources are the papers and datasets the note draws on.
View the original post ↗Embed this note
More notes
The air is now being asked to keep its own ledger
the air is now being asked to keep its own ledger: ecmwf’s aifs compo becomes the first ai model to forecast atmospheric composition globally every three hours, cleanair simulates 365 days of pm2.5 over china in ten seconds, and a unified framework maps six pollutants at one kilometer across the whole country. the air now files its own composition report.
read the note →The current is now being asked to draw its own map
the current is now being asked to draw its own map: china’s langya 2.0 predicts six ocean phenomena including internal waves and mesoscale eddies, a deep net called wenhai resolves eddies globally with air sea flux formulas built in, and scripps infers surface currents from the way temperature patterns deform in satellite images. the ocean now files its own circulation report.
read the note →The soil is now being asked to report its own carbon
the soil is now being asked to report its own carbon: a nix color sensor paired with generative data augmentation predicts soil organic carbon without a lab, random forest drives 74 percent of soil health mapping studies, and sentinel 2 tracks five year carbon change across france and italy from 922 samples. the dirt now files its own carbon account.
read the note →