Liquid AI releases open-weight d1-3B and d1-omni-600M decision models for edge devices
Overview
Liquid AI has released two open-weight decision models, d1-3B and d1-omni-600M (experimental), on Hugging Face, which the company calls early research still under development.
Liquid AI says d1-3B scores 48.57 on the Decision Index v0.2.1, the highest among decision models under 10B parameters; these are the company's own results, not independent benchmarks.
Latency claims vary by source. Liquid AI's first post gave 16 ms on an NVIDIA Jetson AGX Thor and under 50 ms on a Jetson Orin Nano, and its later posts give 8 ms on an RTX 4090, measured one request at a time. MarkTechPost, citing the same index, also reports 8 ms on the RTX 4090 and says the models answer in one forward pass with zero output tokens.
MarkTechPost reports the models are licensed under the LFM Open License v1.0, which allows free commercial use only below $10 million in annual revenue.
Written by AI from the articles below · updated Oct 8, 9:09 PM ET
Check the sources:
Developments
3 developments
- Oct 7, 1:01 PM ET · 1 articleLiquid AI says its open d1 models run across NVIDIA hardware with day-one llama.cpp support and reports d1-3B latencyLiquid AI: Both Open d1 models run across NVIDIA DGX, RTX, and Jetson, with day-one llama.cpp support to run anywhere.
- Oct 7, 1:01 PM ET · 1 articleLiquid AI's d1-3B ranks first among sub-10B models on Decision Index v0.2.1Liquid AI: d1-3B ranks first among models under 10B on the Decision Index v0.2.1, a benchmark for structured decision-making.
- Oct 7, 12:54 PM ET · 4 articlesLiquid AI releases open-weight d1-3B and d1-omni-600M decision modelsHugging Face Blog: Liquid AI Releases Open d1-3B and d1-omni-600M Edge Decision Models
Article timeline
Follow the coverage from different perspectives. Times are ET.
- MarkTechPostLiquid AI releases open-weight d1-3B and d1-omni-600M decision models
AILiquid AI released two open-weight multimodal decision models, d1-3B and d1-omni-600M, which return probability answers in one forward pass with zero output tokens. d1-3B scores 48.57 on Decision Index v0.2.1 and answers one question in 8 ms on an RTX 4090, while the models are licensed free for commercial use below $10 million in annual revenue.
- Liquid AIChoose d1-3B for high-quality text and vision decisions, or explore our experimental d1-omni-600M when footprint matters.
AIDownload the weights, fine-tune, and deploy locally. > Blog: > d1-3B: > d1-omni-600M: > Docs:
- Liquid AIBoth Open d1 models run across NVIDIA DGX, RTX, and Jetson, with day-one llama.cpp support to run anywhere.
AId1-3B single-question latency, measured one request at a time: > NVIDIA RTX 4090: 8 ms > Jetson AGX Thor: 16 ms > Jetson AGX Orin 64 GB: 26 ms > Jetson Orin Nano: 50 ms 4/
- Liquid AId1-3B ranks first among models under 10B on the Decision Index v0.2.1, a benchmark for structured decision-making.
AIBuilt from LFM2.5-VL-3B, it makes decisions from text and images in one pass. Use it for reranking, agent guardrails, and visual inspection. 2/
- Liquid AILiquid AI releases open-weight d1-3B and d1-omni-600M multimodal models
AILiquid AI released Open d1, two open-weight multimodal models in its d1 decision model family. The d1-3B model supports text and vision, while d1-omni-600M supports text plus image or text plus audio. The source says the models are meant for real-time decision making across data centers, RTX workstations, and Jetson edge devices.
- Hugging Face BlogLiquid AI Releases Open d1-3B and d1-omni-600M Edge Decision Models
AILiquid AI released two open-weight decision models, d1-3B and d1-omni-600M (experimental), built on its Liquid Foundation Models and available on Hugging Face. d1-3B scores 48.57 on the Decision Index 0.2.1, the highest among decision models under 10B parameters, and answers a question in 16 ms on an NVIDIA Jetson AGX Thor and under 50 ms on a Jetson Orin Nano. The models support text and images (d1-3B) or text with image or audio (d1-omni-600M).
- Liquid AI BlogPickLiquid AI releases open d1-3B and d1-omni-600M decision models for edge devices
AILiquid AI released two open-weight d1 decision models, d1-3B and d1-omni-600M, on Hugging Face. d1-3B scores 48.57 on the Decision Index v0.2.1 public split and answers a single question in 8 ms on an NVIDIA GeForce RTX 4090 and 50 ms on a Jetson Orin Nano. d1-omni-600M is an experimental checkpoint that handles text with images or audio and scores 15.95 on the same index.
Heat trend
Not enough continuous observations to show a trend yet.