As of: 2026-10-10 23:43 CEST.
Researched and written by AI agents (Claude Opus 5.5 in Claude Code). Human editor: Adam Bicz.
Canonical page: https://postcutoff.com/m/liquid-d1-3b/

# Liquid AI d1-3B

Liquid AI, d1 family. Status: current. Type: language model. Released: 2026-10-07. Training cutoff: not published.

## How to call it

| Provider | Model id | Endpoint | Docs |
|---|---|---|---|
| Hugging Face |  |  | https://huggingface.co/LiquidAI/d1-3B |
| Hugging Face (GGUF) |  |  | https://huggingface.co/LiquidAI/d1-3B-GGUF |
| Hugging Face (d1-omni-600M, text+vision or text+audio) |  |  | https://huggingface.co/LiquidAI/d1-omni-600M |
| Liquid AI Console (hosted d1, $0.04 per 1M input tokens) |  |  | https://console.liquid.ai |

## Pricing

- Price not published.
- Last checked: 2026-10-10

## Specs

- Input: text, image
- Output: text
- Context window: 32,768 tokens
- Max output: not published
- Open weights: yes
- Licence: lfm1.0

## What it has not seen

- No published training cutoff. Counted from its release (2026-10-07): at least 63 AI events since (10 major, 1 historic), as of 2026-10-10.
- Everything since September 2026: https://postcutoff.com/since/2026-09/

## What stands out

- Edge decision model: Returns calibrated typed answers (yes/no, choice, 2-10 level score) in one forward pass with zero output tokens; 8 ms per question on an RTX 4090 and 50 ms on a Jetson Orin Nano (Liquid AI). Source: https://liquid.ai/blog/d1-open

## Notes

Built on LFM2.5-VL-3B (SigLIP2 NaFlex 400M vision encoder). Decision Index v0.2.1 public split 48.57 (self-reported). The hosted d1 (Oct 5) is a separate API model; its size is not stated. Hosted API model id not verified.

Open-weight member of Liquid AI's d1 decision-model family (see the entry for the hosted d1 and d1-omni-600M).

Model page: https://postcutoff.com/m/liquid-d1-3b/