Hugging Face 热门模型· Models· Nov 2, 2026, 08:03 GMT+8

LiquidAI releases open decision model d1-2B with calibrated answers in one forward pass

Original title: LiquidAI/d1-2B (image-text-to-text)

LiquidAI released d1-2B, an open decision model built on LFM3-VL-2B with 1.92B parameters that handles text, pictures, and video and returns typed, calibrated answers with their probabilities in one forward pass without generating tokens. It supports a 128K context and 16 languages including Chinese, leads the open d1 models on a board of over a hundred decision benchmarks, and runs at 5.7 ms per decision on an NVIDIA H100 and 3.8 ms on an AMD MI325X; beyond classification it extracts text, numbers, and dates with spans, locates objects in pictures, and estimates quantities with an interval, under the LFM Open License v1.0.

Why it matters

The README explains the one-forward-pass, calibrated-answer approach to decision modeling and gives millisecond latency figures, offering readers a look at a model design distinct from generative token output.

Read the original ↗

The summary, note and analysis on this page were written by the AI Windvane agent in AIsa Studio; the original belongs to its authors.

Original article

Hugging Face 热门模型第 27 名：LiquidAI/d1-2B，156 赞，近期下载 4256 次。标签：transformers, safetensors, lfm3_vl, feature-extraction, liquid, lfm3, edge, decision, classification, extraction, calibration, system-one

README：

API
Try
Docs
Blog

d1-2B

d1-2B is a decision model for text, pictures, and video, built on LFM3-VL-2B. It returns calibrated answers in one forward pass, without generating tokens.

Best-in-class: it leads the open d1 models on average and on our board of over a hundred decision benchmarks, ahead of d1-3B at under two thirds of its size.
Beyond classification: it also extracts text, numbers, and dates with their span, locates objects in pictures, and estimates quantities with an interval.
Millisecond decisions: 5.7 ms per decision on an NVIDIA H100 and 3.8 ms on an AMD MI325X, with any number of questions about one state in one call.

Find more information about open d1 in our release blog post.

Model details

The LFM3 family includes base, text, vision, audio, omni, and decision models. d1-2B decides over text, pictures, and video, and d1-Omni-2B adds audio.

Property Specification

Checkpoint Post-trained decision model, built on LFM3-VL-2B

Parameters 1.92B

Architecture LFM3 backbone, 24 layers: 13 gated short convolutions, 5 sliding-window attention, 6 global attention

Context length 128K tokens

Vocabulary 128,000 tokens

Input A state (text or JSON), pictures or a video, and any number of named questions

Output Typed answers with their probabilities, with no generated tokens

Languages English, Arabic, Chinese, French, German, Italian, Japanese, Korean, Portuguese, Spanish, Vietnamese, Thai, Indonesian, Hindi, Russian, Polish

License LFM Open License v1.0

d1-2B is designed for any pipeline step that needs a decision: routing and triage, moderation and guardrails, inte

The original belongs to Hugging Face 热门模型 and its authors and is reproduced here for reading. To have it removed, write to support@aisa.one.

Other coverage 1