> ## Documentation Index
> Fetch the complete documentation index at: https://docs.liquid.ai/llms.txt
> Use this file to discover all available pages before exploring further.

# Vision Models

> Liquid's LFM vision models pair our lightweight LFM text backbones with dynamic SigLIP2 image encoders, delivering fast multimodal inference on-device while matching larger VLMs in quality.

<div className="use-cases" style={{marginTop: '-0.5rem'}}>
  <CardGroup cols={2}>
    <Card title="Image Captioning" icon="image">
      Detailed descriptions, alt-text generation, and visual summarization.
    </Card>

    <Card title="OCR & Extraction" icon="file-lines">
      Text recognition, form parsing, and document digitization.
    </Card>

    <Card title="Visual Reasoning" icon="lightbulb">
      Scene understanding, spatial relations, and visual Q\&A.
    </Card>

    <Card title="Continuous Capture" icon="video">
      Always-on activity recognition and scene monitoring on-device.
    </Card>
  </CardGroup>
</div>

<h2>LFM2.5 Models   <Badge shape="rounded" icon="circle-check" color="green" size="lg">Latest release</Badge></h2>

<Info>LFM2.5-VL builds on LFM2-VL with extended reinforcment learning training for higher performance while maintaining the same architecture and deployment footprint.</Info>

<CardGroup cols={2}>
  <Card title="LFM2.5-VL-1.6B" href="/lfm/models/lfm25-vl-1.6b">
    1.6B · <Badge shape="pill" color="green">Recommended</Badge>

    Best vision model for most use cases. Fast and accurate.
  </Card>

  <Card title="LFM2.5-VL-450M" href="/lfm/models/lfm25-vl-450m">
    450M · <Badge shape="pill" color="yellow">Fastest</Badge>

    Compact vision model for edge deployment and fast inference.
  </Card>
</CardGroup>

## LFM2 Models

<CardGroup cols={3}>
  <Card title="LFM2-VL-3B" href="/lfm/models/lfm2-vl-3b">
    3B

    Highest-capacity multimodal model with enhanced visual reasoning.
  </Card>

  <Card title="LFM2-VL-1.6B" href="/lfm/models/lfm2-vl-1.6b">
    1.6B · <Badge shape="pill" color="gray">Deprecated</Badge>

    Use the new LFM2.5-VL-1.6B checkpoint instead.
  </Card>

  <Card title="LFM2-VL-450M" href="/lfm/models/lfm2-vl-450m">
    450M · <Badge shape="pill" color="yellow">Fastest</Badge>

    Compact multimodal model for edge deployment and fast inference.
  </Card>
</CardGroup>

## Examples

Explore practical implementations using vision models:

<CardGroup cols={2}>
  <Card title="Image Understanding with Vision Language Models" href="/examples/android/vision-language-model-example">
    Analyze images, answer visual questions, and generate descriptions using LFM2-VL-1.6B on Android with Jetpack Compose and Coil.

    **Platform:** Android · **Uses:** LFM2-VL-1.6B
  </Card>

  <Card title="Invoice Extractor Tool" href="/examples/laptop-examples/invoice-extractor-tool-with-liquid-nanos">
    Extract structured payment data from invoice PDFs using LFM2.5-VL-1.6B with file monitoring and 100% local processing.

    **Platform:** Desktop · **Uses:** LFM2.5-VL-1.6B
  </Card>

  <Card title="Real-Time Video Captioning" href="/examples/web/vl-webgpu-demo">
    Generate video captions directly in-browser using LFM2.5-VL-1.6B with WebGPU acceleration and ONNX Runtime Web.

    **Platform:** Web · **Uses:** LFM2.5-VL-1.6B
  </Card>

  <Card title="Fine-tune for Car Maker Identification" href="/examples/customize-models/car-maker-identification">
    Learn to fine-tune LFM2-VL models (450M, 1.6B, 3B) with LoRA, structured generation, and evaluation pipelines for image classification.

    **Platform:** Desktop · **Uses:** LFM2-VL-450M, 1.6B, 3B
  </Card>
</CardGroup>

<Panel>
  * [Liquid Playground](https://playground.liquid.ai/chat?model=cmk0wefde000204jp2knb2qr8)
  * [HuggingFace Collections](https://huggingface.co/LiquidAI/collections)
  * [OpenRouter API](https://openrouter.ai/liquid)
  * [LEAP Model Library](https://leap.liquid.ai/models)
</Panel>