---
title: "Together AI: Fast Open-Source LLM Inference with Llama 3.3 and FLUX in Pixeltable"
date: "2025-04-10"
author: "Pixeltable Team"
tags:
  - Together AI
  - Llama 3.3
  - Mixtral
  - FLUX
  - Open Source LLM
  - AI Integration
  - Pixeltable
description: "Build high-performance AI applications with Together AI's optimized inference platform. Learn how to use Llama 3.3, Mixtral, FLUX, and other open-source models with Pixeltable for production workloads."
url: "https://pixeltable.com/blog/together-ai-llm-inference-pixeltable"
---

# Together AI: Fast Open-Source LLM Inference with Llama 3.3 and FLUX in Pixeltable

Together AI delivers fast, reliable inference for open-source models with competitive pricing. Combined with Pixeltable, you can build production AI applications that leverage the best open-source models.

 
## Together AI: Open-Source Model Specialists

 
**Together AI** has built a world-class inference platform optimized for open-source models. Their infrastructure delivers fast responses with high reliability.

 
 
When combined with Pixeltable's [declarative infrastructure](/blog/unified-multimodal-ai-infrastructure-pixeltable), you get reliable inference with enterprise-grade orchestration.

 
## Available Models

 

 - **Llama 3.3 70B Instruct Turbo:** Fastest Llama inference

 - **Mixtral 8x7B:** Efficient mixture of experts

 - **Qwen 2.5:** Multilingual excellence

 - **FLUX Schnell:** Fast image generation

 - **BGE Embeddings:** High-quality vectors

 

 
## Getting Started

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

# Set your Together AI API key
export TOGETHER_API_KEY="your-api-key-here"
```

 
## Chat Completions

 
```python
import pixeltable as pxt
from pixeltable.functions import together

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

t = pxt.create_table('together_demo.chat', {'input': pxt.String})

messages = [{'role': 'user', 'content': t.input}]
t.add_computed_column(output=together.chat_completions(
 messages=messages,
 model='meta-llama/Llama-3.3-70B-Instruct-Turbo',
 model_kwargs={'max_tokens': 300}
))

t.add_computed_column(response=t.output.choices[0].message.content)

t.insert([{'input': 'What is the capital of France?'}])
t.select(t.input, t.response).show()
```

 
## Embeddings with BGE

 
```python
import pixeltable as pxt
from pixeltable.functions import together

docs = pxt.create_table('together_demo.docs', {'text': pxt.String})

docs.add_computed_column(
 embedding=together.embeddings(
 input=docs.text,
 model='BAAI/bge-large-en-v1.5'
 )
)

docs.insert([
 {'text': 'Machine learning transforms industries'},
 {'text': 'Deep learning is a subset of ML'}
])

docs.show()
```

 
## Image Generation with FLUX

 
```python
import pixeltable as pxt
from pixeltable.functions import together

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

images.add_computed_column(
 generated=together.image_generation(
 prompt=images.prompt,
 model='black-forest-labs/FLUX.1-schnell-Free'
 )
)

images.insert([{'prompt': 'A serene mountain landscape at sunset'}])
images.show()
```

 
## Pricing Overview

 
| Model | Input (per 1M tokens) | Output (per 1M tokens) |
| --- | --- | --- |
| Llama 3.3 70B Turbo | $0.88 | $0.88 |
| Mixtral 8x7B | $0.60 | $0.60 |
| BGE Large | $0.02 | - |

 
## Together AI vs Other Providers

 
| Feature | Together AI | Fireworks | Groq |
| --- | --- | --- | --- |
| Speed | Fast | Fast | Fastest |
| Model Selection | Extensive | Extensive | Limited |
| Image Generation | Yes (FLUX) | Limited | No |
| Embeddings | Yes | Limited | No |

 
## Next Steps

 

 - [Try Groq for fastest inference](/blog/groq-fast-inference-pixeltable)

 - [Production RAG Guide](/blog/production-rag-data-centric)

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

 

 
## Resources

 

 - [Pixeltable Together AI Documentation](https://docs.pixeltable.com/howto/providers/working-with-together)

 - [Together AI Documentation](https://docs.together.ai)

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