Bringing Secure AI to Your Workflow: Deepseek R1 and Llama Now Available in Storytell

Storytell now brings Deepseek and Llama models into your workflow, combining cutting-edge capabilities with enterprise-grade protection. Whether you're tackling complex reasoning, debugging code, or generating high-quality content, these models are now at your command. Just type the keywords “use Deepseek” or “use Llama” along with your prompt in SmartChat™ and let Storytell select the model for you.

With secure US-based hosting and compliance built in, you get the power of AI without the risks. Ready to put these models to work? Here’s how.

Why this matters

You've asked for powerful AI capabilities that don't compromise on security. By hosting Llama and a distilled version of Deepseek on US-based servers, we're providing you with cutting-edge AI technology while ensuring complete compliance with data privacy regulations. This means you can leverage advanced AI features with confidence, knowing your data never leaves secure US infrastructure.

Key features you can start using today

Enterprise-safe model integration

Both models are now available through our platform with a simple command structure:

• Access Deepseek by appending "Use Deepseek" to your queries

• Access Llama by adding "Use Llama" to your prompts

Specialized capabilities for different needs

Deepseek powers:

• Complex reasoning tasks
• Advanced computational challenges
• Structured problem-solving scenarios

Llama excels at:

• Natural language understanding
• Enhanced text generation
• Versatile communication tasks

Security and Compliance features

• US-based server hosting
• Enterprise-grade security protocols
• Full compliance with data privacy regulations
• Protected data handling without external exposure

How to get started

Using these new models is straightforward:

1. Choose your preferred model based on your task requirements
2. Append the appropriate command to your query on SmartChat™
3. Get results that leverage each model's unique strengths

Who benefits from this update?

Enterprise teams

With US-based infrastructure, sensitive data remains protected while you utilize powerful AI features.

Developers and engineers

Benefit from enhanced support for:

• Advanced coding assistance
• Debugging support
• Structured problem-solving tools

Researchers

Access powerful tools for:

• Logic-driven query processing
• Complex computational tasks
• Advanced analytical processes

Current implementation status

While both models operate as standalone integrations within our ecosystem, they're not yet embedded in the LLM router. This means you'll need to manually select your preferred model for specific tasks. We've designed this implementation to allow you to test and evaluate each model's strengths for your unique use cases.

Technical note on model architecture

You might notice when using Deepseek for your queries that it will display “Answered by DeepSeek R1 Distill Llama 70B.” Why?

For those interested in the technical details, this model is a fine-tuned version of the Llama 3.3 70B model. This means the base Llama model has been trained further using samples generated by DeepSeek, enhancing its capabilities in tasks like reasoning and understanding specific types of data. The "70b" refers to the model's size, with 70 billion parameters, making it a large and powerful tool for complex tasks.

Here's a simpler explanation: The DeepSeek R1 Distill Llama 70b model starts with a powerful base model called Llama. We then take another advanced model, DeepSeek, and use its expertise to further train Llama, making it even better at understanding and solving complex problems. The '70b' means it has 70 billion 'brain cells,' allowing it to handle really tough tasks. This fine-tuning process is like giving Llama specialized lessons to enhance its skills in specific areas, making it more powerful and effective for your needs.

Looking forward

This integration marks just the beginning of our commitment to providing secure, powerful AI solutions. We're actively working on enhancing the integration and accessibility of these models to make your experience even better.

Important notes about model selection

To make the most of these new capabilities:

• Consider using Deepseek for tasks requiring complex reasoning or computational work
• Choose Llama when you need robust natural language understanding or text generation
• Remember to explicitly select your model using the appropriate command

We're proud to offer you this enterprise-safe integration of Deepseek and Llama models. By hosting these powerful tools on secure US-based infrastructure, we're ensuring you can leverage cutting-edge AI while maintaining the highest standards of data protection and compliance.

Ready to explore these new capabilities? Try it on Storytell.ai today.