Text / LiquidAI

LFM2.5 VL 3B

LFM2.5-VL-3B is a multimodal variant of the LFM2.5 hybrid model family designed for on-device deployment. It processes text and images, using LFM2.5-2.6B as its language backbone and a SigLIP2 NaFlex vision encoder. The model reports improved grounding and object detection with natural language queries, full-page OCR with layout annotation, and efficient inference (reported 228 tok/s on Apple M5 Max and 116 tok/s on AMD Ryzen AI Max+ 395, under 3.3 GB memory). It supports 16 languages and a 32,768-token context length, with native resolution processing that splits large images into non-overlapping 512x512 patches plus a resized whole-image thumbnail. The model is available under the LFM 1.0 license and is tagged for image-text-to-text, with exports including GGUF, ONNX, and MLX variants.

EstimatedRepository opened Aug 11, 2026Source checked 9/24/2026Version: a3af5799

LOCALRENTED GPUOPEN WEIGHTS

Freshness 78 out of 100. Repository age: 65 days old, Aug 11, 2026. Measures age, not quality.

Popularity 60 out of 100. 29,488 downloads in 30 days; 225 lifetime likes. Checked Oct 15, 2026. Measures adoption, not quality.

What do these scores mean?

Two ways to compare models. Neither score measures answer quality, benchmark performance, or whether a model fits your machine.

Freshness · release age

A new release starts at 100. The score halves every 180 days: 100 × 2^(−age in days / 180). When the release date is unavailable, a repository date is a fallback—not proof of when the model was released. Missing dates stay unscored.

Check the date source

Popularity · adoption

Downloads over the past 30 days carry 80% of the score; lifetime likes carry 20%. Both use a logarithmic scale, capped at 10 million downloads and 100,000 likes. If likes are unavailable, downloads carry all the weight. Missing download statistics stay unscored.

round(80 × min(1, log10(1 + downloads) / 7) + 20 × min(1, log10(1 + likes) / 5))

Check the statistics source

Scores as of Oct 15, 2026. Statistics checked Oct 15, 2026. Statistics are marked for refresh after 72 hours.Full scoring method

At a glance

Parameters: 3.12B
Architecture: lfm2_vl
License: lfm1.0
Disk space: 5.8 GB
Software: Transformers

View the model source

MY HARDWARE

Will it run on your machine?

Save your machine to see a personalized rating and its reasoning.

Add my hardware

Best for

Image understanding and visual question answering with text and images
Grounding and object detection using natural language queries
Full-page optical character recognition (OCR) with layout annotation
Multilingual multimodal interactions across 16 supported languages
On-device deployment scenarios with native, GGUF, ONNX, or MLX formats

Tradeoffs

License is LFM 1.0 (other), not an open-source or permissive license
The source does not specify hardware requirements beyond reported inference metrics; no compatibility or performance guarantees are provided
No claims of testing or evaluation are made in the supplied data
The source does not include information on fine-tuning support, quantization details, or numerical accuracy
No recommendations for purchase or comparisons with other models are provided in the source content

Ways to run it

Transformers · See official guide

Repository-specific command in publisher documentation; confirm dependencies and hardware in the source.

import torch
from transformers import AutoModelForImageTextToText, AutoProcessor

model_id = "LiquidAI/LFM2.5-VL-3B"

model = AutoModelForImageTextToText.from_pretrained(model_id, dtype=torch.bfloat16)
processor = AutoProcessor.from_pretrained(model_id)

messages = [
    {
        "role": "user",
        "content": [
            {"type": "image", "url": "https://placecats.com/300/200"},
            {"type": "text", "text": "Describe this image."},
        ],
    }
]

inputs = processor.apply_chat_template(
    messages,
    add_generation_prompt=True,
    tokenize=True,
    return_dict=True,
    return_tensors="pt",
).to(model.device)

with torch.inference_mode():
    output_ids = model.generate(
        **inputs,
        do_sample=True,
        temperature=0.2,
        top_k=50,
        repetition_penalty=1.0,
        max_new_tokens=256,
    )

generated_ids = output_ids[:, inputs["input_ids"].shape[1] :]
print(processor.batch_decode(generated_ids, skip_special_tokens=True)[0])

Official instructions

Explore its uses

SAFETENSORSIMAGE-TEXT-TO-TEXT

Keep exploring

Text

Prism ML

Bonsai 2 · 27B

A compressed 27B-class reasoning model with publisher-provided GGUF packs and a dedicated llama.cpp fork for CUDA, Metal, and CPU.

Freshness 90 out of 100. Release age: 28 days old, Sep 17, 2026. Measures age, not quality.

MODEL SIZE27.36B

LOCALRENTED GPU

Released Sep 17, 2026

Text

DeepSeek

DeepSeek V4.1 Flash

A multimodal reasoning model with a compressed key-value cache, published as open weights.

Freshness 87 out of 100. Repository age: 35 days old, Sep 10, 2026. Measures age, not quality.

Arena Open Source Text: rank 5 of 225. Ranked model: deepseek-v4.1-flash-max. Export dated Oct 13, 2026. Source checked Oct 15, 2026. Human preference ranking, not hardware suitability.+3

LOCALRENTED GPU

Repository opened Sep 10, 2026

Text

Qwen

Qwen3.8

Explore 4 versions of Qwen3.8. Compare model sizes, packages, hardware requirements, and available setups.

4 versionsCompare sizes & formats

LOCALRENTED GPU

Compare versions →