> ## Documentation Index
> Fetch the complete documentation index at: https://mintlify.com/mlfoundations/open_clip/llms.txt
> Use this file to discover all available pages before exploring further.

# Pretrained Models

> Complete catalog of pretrained OpenCLIP models with performance metrics, training details, and usage examples

OpenCLIP provides a comprehensive collection of pretrained CLIP models trained on various datasets at different scales.

## Listing Available Models

### List All Pretrained Models

```python theme={null}
import open_clip

# Get all pretrained model/tag combinations
pretrained = open_clip.list_pretrained()
for model_name, tag in pretrained:
    print(f"{model_name}:{tag}")

# Get as formatted strings
pretrained_str = open_clip.list_pretrained(as_str=True)
# Returns: ['RN50:openai', 'ViT-B-32:laion2b_s34b_b79k', ...]
```

### List Model Architectures

```python theme={null}
import open_clip

# Get all available model architectures
architectures = open_clip.list_models()
print(architectures)
# ['RN50', 'RN101', 'RN50x4', 'ViT-B-32', 'ViT-L-14', ...]
```

### Query Specific Models

```python theme={null}
from open_clip import list_pretrained_tags_by_model, list_pretrained_models_by_tag

# Get all pretrained tags for a specific model
tags = list_pretrained_tags_by_model('ViT-B-32')
print(tags)
# ['openai', 'laion400m_e31', 'laion2b_s34b_b79k', 'datacomp_xl_s13b_b90k', ...]

# Get all models with a specific tag
models = list_pretrained_models_by_tag('openai')
print(models)
# ['RN50', 'RN101', 'ViT-B-32', 'ViT-L-14', ...]
```

## Model Zoo Overview

### Vision Transformer (ViT) Models

ViT models provide excellent accuracy-efficiency tradeoffs:

| Model       | Parameters | FLOPs  | Best Pretrained Tag      | ImageNet Zero-Shot |
| ----------- | ---------- | ------ | ------------------------ | ------------------ |
| ViT-B-32    | 88M        | 4.4G   | datacomp\_xl\_s13b\_b90k | 72.8%              |
| ViT-B-16    | 86M        | 17.6G  | datacomp\_xl\_s13b\_b90k | 73.5%              |
| ViT-L-14    | 304M       | 81.1G  | datacomp\_xl\_s13b\_b90k | 79.2%              |
| ViT-H-14    | 632M       | 169.1G | laion2b\_s32b\_b79k      | 78.0%              |
| ViT-g-14    | 1.01B      | 257.4G | laion2b\_s34b\_b88k      | 76.6%              |
| ViT-bigG-14 | 1.84B      | 571.3G | laion2b\_s39b\_b160k     | 80.1%              |

### ResNet Models

Classic CNN-based architectures:

| Model   | Parameters | ImageNet Zero-Shot | Best Tag |
| ------- | ---------- | ------------------ | -------- |
| RN50    | 38M        | 59.6%              | openai   |
| RN101   | 56M        | 62.3%              | openai   |
| RN50x4  | 87M        | 68.3%              | openai   |
| RN50x16 | 167M       | 73.2%              | openai   |
| RN50x64 | 623M       | 76.6%              | openai   |

### ConvNeXt Models

Modern CNN architectures with competitive performance:

| Model            | Parameters | ImageNet Zero-Shot | Best Tag                    |
| ---------------- | ---------- | ------------------ | --------------------------- |
| ConvNeXt-Base    | 88M        | 71.5%              | laion2b\_s13b\_b82k\_augreg |
| ConvNeXt-Base-W  | 89M        | 72.1%              | laion2b\_s13b\_b82k\_augreg |
| ConvNeXt-Large-D | 200M       | 76.9%              | laion2b\_s29b\_b131k\_ft    |
| ConvNeXt-XXLarge | 846M       | 79.5%              | laion2b\_s34b\_b82k\_augreg |

### SigLIP Models

Models trained with SigLIP loss function:

| Model                | Resolution | ImageNet Zero-Shot | Tag   |
| -------------------- | ---------- | ------------------ | ----- |
| ViT-B-16-SigLIP      | 256        | 76.9%              | webli |
| ViT-B-16-SigLIP      | 384        | 78.4%              | webli |
| ViT-L-16-SigLIP      | 256        | 79.8%              | webli |
| ViT-L-16-SigLIP      | 384        | 81.5%              | webli |
| ViT-SO400M-14-SigLIP | 224        | 82.0%              | webli |
| ViT-SO400M-14-SigLIP | 384        | 83.1%              | webli |

### EVA Models

State-of-the-art EVA-CLIP models:

| Model           | ImageNet Zero-Shot | Tag                  |
| --------------- | ------------------ | -------------------- |
| EVA02-B-16      | 74.7%              | merged2b\_s8b\_b131k |
| EVA02-L-14      | 79.8%              | merged2b\_s4b\_b131k |
| EVA02-L-14-336  | 80.4%              | merged2b\_s6b\_b61k  |
| EVA02-E-14      | 81.9%              | laion2b\_s4b\_b115k  |
| EVA02-E-14-plus | 82.7%              | laion2b\_s9b\_b144k  |

## Training Dataset Information

### OpenAI WIT

```python theme={null}
model, _, preprocess = open_clip.create_model_and_transforms(
    'ViT-B-32', 
    pretrained='openai'
)
```

* **Dataset**: WebImageText (WIT) - 400M image-text pairs
* **Models**: Original OpenAI CLIP models
* **Tags**: `openai`

### LAION Datasets

```python theme={null}
model, _, preprocess = open_clip.create_model_and_transforms(
    'ViT-L-14',
    pretrained='laion2b_s32b_b82k'
)
```

* **LAION-400M**: 400M English image-text pairs
* **LAION-2B**: 2B English image-text pairs
* **LAION-5B**: 5B multilingual image-text pairs
* **Tags**: `laion400m_*`, `laion2b_*`, `laion5b_*`

### DataComp

```python theme={null}
model, _, preprocess = open_clip.create_model_and_transforms(
    'ViT-L-14',
    pretrained='datacomp_xl_s13b_b90k'
)
```

* **DataComp-XL**: 1.4B filtered image-text pairs
* **DataComp-L**: 1.0B pairs
* **DataComp-M**: 128M pairs
* **DataComp-S**: 13M pairs
* **Tags**: `datacomp_xl_*`, `datacomp_l_*`, `datacomp_m_*`, `datacomp_s_*`

### WebLI (SigLIP)

```python theme={null}
model, _, preprocess = open_clip.create_model_and_transforms(
    'ViT-SO400M-14-SigLIP-384',
    pretrained='webli'
)
```

* **Dataset**: Google's WebLI dataset
* **Models**: SigLIP models with sigmoid loss
* **Tags**: `webli`

## Model Selection Guide

<Tabs>
  <Tab title="Speed Priority">
    **Fastest inference:**

    ```python theme={null}
    # Small and fast
    model, _, preprocess = open_clip.create_model_and_transforms(
        'ViT-B-32',  # 4.4 GFLOPs
        pretrained='laion2b_s34b_b79k'
    )
    ```

    * **ViT-B-32**: Best speed/accuracy tradeoff
    * **RN50**: Classic CNN option
    * **MobileCLIP-S1**: Ultra-fast mobile deployment
  </Tab>

  <Tab title="Accuracy Priority">
    **Best zero-shot accuracy:**

    ```python theme={null}
    # State-of-the-art accuracy
    model, _, preprocess = open_clip.create_model_and_transforms(
        'ViT-bigG-14',  # 80.1% ImageNet
        pretrained='laion2b_s39b_b160k'
    )
    ```

    * **ViT-bigG-14**: 80.1% ImageNet zero-shot
    * **ViT-SO400M-14-SigLIP-384**: 83.1% (SigLIP)
    * **EVA02-E-14-plus**: 82.7% (EVA-CLIP)
  </Tab>

  <Tab title="Balanced">
    **Good balance of speed and accuracy:**

    ```python theme={null}
    # Recommended for most use cases
    model, _, preprocess = open_clip.create_model_and_transforms(
        'ViT-L-14',  # 79.2% ImageNet
        pretrained='datacomp_xl_s13b_b90k'
    )
    ```

    * **ViT-L-14**: Excellent performance at reasonable cost
    * **ViT-B-16**: Good accuracy, moderate speed
    * **ConvNeXt-Base**: CNN alternative
  </Tab>
</Tabs>

## Loading from HuggingFace Hub

Many models are available on HuggingFace Hub:

```python theme={null}
import open_clip

# Direct HuggingFace loading
model, _, preprocess = open_clip.create_model_and_transforms(
    'hf-hub:laion/CLIP-ViT-L-14-DataComp.XL-s13B-b90K',
    device='cuda'
)

# Alternative: use standard loading
model, _, preprocess = open_clip.create_model_and_transforms(
    'ViT-L-14',
    pretrained='datacomp_xl_s13b_b90k',
    device='cuda'
)
```

### Popular HuggingFace Models

* `laion/CLIP-ViT-L-14-DataComp.XL-s13B-b90K`
* `laion/CLIP-ViT-H-14-laion2B-s32B-b79K`
* `laion/CLIP-ViT-bigG-14-laion2B-39B-b160k`
* `apple/DFN5B-CLIP-ViT-H-14`
* `timm/ViT-SO400M-14-SigLIP-384`

<Note>
  HuggingFace models automatically download config and weights. The `hf-hub:` prefix explicitly specifies HuggingFace as the source.
</Note>

## Multilingual Models

Models trained on multilingual datasets:

```python theme={null}
# XLM-RoBERTa based models
model, _, preprocess = open_clip.create_model_and_transforms(
    'xlm-roberta-base-ViT-B-32',
    pretrained='laion5b_s13b_b90k',
    device='cuda'
)

# Supports 100+ languages
tokenizer = open_clip.get_tokenizer('xlm-roberta-base-ViT-B-32')
```

Available multilingual models:

* `xlm-roberta-base-ViT-B-32`
* `xlm-roberta-large-ViT-H-14`
* `ViT-B-16-SigLIP-i18n-256` (SigLIP multilingual)

## Specialized Models

### CoCa (Captioning)

Models with generative capabilities:

```python theme={null}
model, _, preprocess = open_clip.create_model_and_transforms(
    'coca_ViT-L-14',
    pretrained='mscoco_finetuned_laion2b_s13b_b90k'
)

# Generate captions
import torch
from PIL import Image

im = preprocess(Image.open("cat.jpg")).unsqueeze(0)
with torch.no_grad(), torch.cuda.amp.autocast():
    generated = model.generate(im)
print(open_clip.decode(generated[0]))
```

### DFN Models

Apple's Distilled Feature Networks:

```python theme={null}
model, _, preprocess = open_clip.create_model_and_transforms(
    'ViT-H-14',
    pretrained='dfn5b',  # 83.4% ImageNet
    device='cuda'
)
```

## Performance Metrics

For detailed zero-shot performance across 38 datasets, see the [full results CSV](https://github.com/mlfoundations/open_clip/blob/main/docs/openclip_results.csv).

### Benchmark Datasets

* **ImageNet**: Primary zero-shot benchmark
* **ImageNet variants**: -A, -R, -Sketch, -V2
* **Object recognition**: CIFAR-10/100, STL-10, Food101
* **Fine-grained**: Flowers102, Pets, Cars, Aircraft
* **Scene**: SUN397, Places365
* **Action**: UCF101, Kinetics700

## Example: Finding Best Model

```python theme={null}
import open_clip

# Find all ViT-L-14 pretrained versions
tags = open_clip.list_pretrained_tags_by_model('ViT-L-14')
print(f"Available ViT-L-14 models: {tags}")

# Load the best performing one
model, _, preprocess = open_clip.create_model_and_transforms(
    'ViT-L-14',
    pretrained='datacomp_xl_s13b_b90k',  # 79.2% ImageNet
    device='cuda',
    precision='fp16'
)

model.eval()
print(f"Model loaded with {sum(p.numel() for p in model.parameters()) / 1e6:.1f}M parameters")
```