Loading...
From lightning-fast responses to complex reasoning, our models are built to handle any AI workload with enterprise-grade reliability.
Our compressed and optimized open-source model available on Hugging Face. Lightning-fast and perfect for real-time applications.
Our compressed and optimized 2B parameter language model, open-sourced on Hugging Face alongside Rax 4.0. Process text with up to 262K context.
Compare our models side-by-side to find the perfect fit for your application.
| Feature | Rax 4.0 | Rax 4.5 |
|---|---|---|
| Context Window | 8,192 tokens | 262,144 tokens |
| Model Parameters | 1.5 Billion | 2 Billion |
| Inference Speed | Very Fast | Fast |
| VRAM Efficiency | Standard | Reduced KV Cache Footprint |
| Streaming Support | ✓ | ✓ |
| License | Apache 2.0 | Apache 2.0 |
Get the best results by choosing the right model for your specific task.
For speed-critical applications
For long-context and complex reasoning tasks
Try both models in our interactive playground or integrate them directly into your application.