Models
Rustellar provides multiple AI models, each optimized for specific use cases.
Available Models
Helix v1
Architecture: Transformer Decoder-only Specialization: General conversation
Helix v1 is our flagship conversational AI model, designed for a wide range of dialogue-based applications. Built on the proven Transformer decoder-only architecture, it excels at understanding context and generating natural, coherent responses.
Iroha
Architecture: Mamba Specialization: Story generation
Iroha is a specialized model built on the Mamba architecture, specifically tuned for creative storytelling. Whether you're generating narratives, character dialogues, or plot developments, Iroha delivers engaging and contextually rich content.
Choosing the Right Model
| Use Case | Recommended Model |
|---|---|
| General conversation & Q&A | Helix v1 |
| Creative writing & storytelling | Iroha |
| Customer support chatbots | Helix v1 |
| Novel or script generation | Iroha |
Model Comparison
| Feature | Helix v1 | Iroha |
|---|---|---|
| Architecture | Transformer Decoder | Mamba |
| Primary Use | Conversation | Story Generation |
| Context Window | 8,192 tokens | 16,384 tokens |
| Response Speed | Fast | Medium |
For detailed specifications and usage examples, visit each model's dedicated page.