---
title: "Replicate: Access Thousands of ML Models for LLMs, Images, and Audio in Pixeltable"
date: "2025-10-15"
author: "Pixeltable Team"
tags:
  - Replicate
  - Model Marketplace
  - Llama
  - FLUX
  - Whisper
  - AI Integration
  - Pixeltable
description: "Build AI applications with Replicate's vast model marketplace. Learn how to use Llama, FLUX, Whisper, and thousands of other models with Pixeltable for production-ready ML pipelines."
url: "https://pixeltable.com/blog/replicate-model-marketplace-pixeltable"
---

# Replicate: Access Thousands of ML Models for LLMs, Images, and Audio in Pixeltable

Replicate provides access to thousands of ML models through a simple API. Combined with Pixeltable, you can build sophisticated multimodal pipelines that leverage the best models for each task.

 
## Replicate: The Model Marketplace

 
**Replicate** is a platform that lets you run open-source machine learning models in the cloud. With thousands of models available, you can build virtually any AI application without managing infrastructure.

 
 
When combined with Pixeltable's [declarative infrastructure](/blog/unified-multimodal-ai-infrastructure-pixeltable), you get access to the entire ML ecosystem with automatic orchestration.

 
## Popular Model Categories

 
 
### Language Models

 

 - **Meta Llama 3:** Latest open-weight LLM family

 - **Mistral/Mixtral:** Efficient European models

 

 
### Image Generation

 

 - **FLUX:** Black Forest Labs' fast image generator

 - **Stable Diffusion XL:** Stability AI's flagship model

 

 
### Audio Processing

 

 - **Whisper:** OpenAI's speech recognition

 - **MusicGen:** Meta's music generation

 

 
## Getting Started

 
```bash
# Install required packages
pip install pixeltable replicate

# Set your Replicate API token
export REPLICATE_API_TOKEN="your-token-here"
```

 
## Using LLMs with Replicate

 
```python
import pixeltable as pxt
from pixeltable.functions.replicate import run

pxt.drop_dir('replicate_demo', force=True)
pxt.create_dir('replicate_demo')

t = pxt.create_table('replicate_demo.chat', {'prompt': pxt.String})

input_config = {
 'system_prompt': 'You are a helpful assistant.',
 'prompt': t.prompt,
 'max_tokens': 300
}

t.add_computed_column(output=run(input_config, ref='meta/meta-llama-3-8b-instruct'))

t.add_computed_column(response=pxt.functions.string.join('', t.output))

t.insert([{'prompt': 'What foods are rich in selenium?'}])
t.select(t.prompt, t.response).show()
```

 
## Image Generation with FLUX

 
```python
import pixeltable as pxt
from pixeltable.functions.replicate import run

images = pxt.create_table('replicate_demo.images', {'prompt': pxt.String})

input_config = {
 'prompt': images.prompt,
 'go_fast': True,
 'megapixels': '1'
}

images.add_computed_column(output=run(input_config, ref='black-forest-labs/flux-schnell'))

images.add_computed_column(image=images.output[0].astype(pxt.Image))

images.insert([
 {'prompt': 'A friendly dinosaur playing tennis'},
 {'prompt': 'A futuristic cityscape at sunset'}
])

images.select(images.prompt, images.image).show()
```

 
## Popular Replicate Models

 
| Category | Model | Reference |
| --- | --- | --- |
| LLM | Llama 3.1 8B | meta/meta-llama-3-8b-instruct |
| Image | FLUX Schnell | black-forest-labs/flux-schnell |
| Audio | Whisper | openai/whisper |
| Video | Stable Video | stability-ai/stable-video-diffusion |

 
## Pricing Model

 

 - **CPU Models:** ~$0.000100 per second

 - **GPU Models:** ~$0.000725 per second

 - **No minimums:** Pay only for what you use

 

 
## Next Steps

 

 - [fal.ai for faster images](/blog/fal-ai-image-generation-pixeltable)

 - [Multi-Provider Strategy](/blog/multi-provider-ai-strategy-comparison-guide)

 - [All Integrations](https://docs.pixeltable.com/integrations/frameworks)

 

 
## Resources

 

 - [Pixeltable Replicate Documentation](https://docs.pixeltable.com/howto/providers/working-with-replicate)

 - [Replicate Documentation](https://replicate.com/docs)

 - [Join our Discord Community](https://discord.com/invite/QPyqFYx2UN)