How to Use DeepSeek-R1 and Llama

NiceGPU Platform provide the quickly use of models such as Llama and DeepSeek-R1 through Ollama, a tool focused on local deployment of large language models.
Currently, multiple large language models such as Llama, Qwen, DeepSeek R1, Phi, and Gemma3 are provided for quick experimentation.

1. Create an Instance

Larger models require more video memory. Please refer to the following table to select a suitable graphics card machine to create an instance.

Template Name	Ollama Restful API model name	Graphics Card Requirement (VRAM)	Recommended Graphics Card Example
Llama 3B	llama3.2	At least 4GB	NVIDIA GTX 1650
EXAONE Deep 7.8B	exaone-deep:7.8b	At least 6GB	NVIDIA RTX 2070 Super
DeepSeek R1 8B	deepseek-r1:8b	At least 6GB	NVIDIA RTX 2070 Super
Llama 11B	llama3.2-vision	At least 12GB	NVIDIA RTX 3090 or higher
DeepSeek R1 14B	deepseek-r1:14b	At least 16GB	NVIDIA RTX 4070S or higher
Gemma3 12B	gemma3:12b	At least 16GB	NVIDIA RTX 4070S or higher
Phi-4 14B	phi4:14b	At least 16GB	NVIDIA RTX 4070S or higher
DeepSeek Coder V2 16B	deepseek-coder-v2:16b	At least 24GB	NVIDIA RTX 4090 or higher

2. Confirm the Running Status

After the instance is successfully launched, you can directly access the service by clicking the Compute Connection button.

You can see that Ollama is in the running state.

3. Confirm the Models

Access /api/tags to confirm the list of currently supported models.

4. Usage Methods

a. Usage with ChatBox or Open Web UI

1. Download a popular LLM Web UI application or opt for its web-based version.
2. In the settings, choose OLLAMA API as the service provider and input your instance address.
3. Start enjoying the experience!

ChatboxaiOpenWeb UI

Click https://web.chatboxai.app to quickly try the Chatbox Web version:

Install the OpenWeb UI program on your own. In the settings, select OLLAMA API as the provider and enter your instance address.

b. Direct Call via API

For more Ollama RESTful API interfaces and parameters, please refer to the official REST API documentation.

Ollama 3BDeepSeek R1 7B

{
    "model": "llama3.2",
    "prompt": "Why is the sky blue?",
    "stream": false
}

{
    "model": "deepseek-r1:7b",
    "prompt": "Why is the sky blue?",
    "stream": false
}