DeepSeek Setup Guide

There are a few ways to use DeepSeek models in 16x Prompt:

1. (Recommended) Connect to DeepSeek API directly (fastest and cheapest)
2. Connect via Hyperbolic (relatively fast and cheap, but can cause quality degradation)
3. Connect via Fireworks (relatively fast, but expensive)
4. Connect via OpenRouter API (slower and potentially more expensive)
5. Connect via DeepInfra (slower, but cheap)
6. Connect via Ollama (slowest, but can be used on local machine)

Read our blog post DeepSeek R1: Comparing Pricing and Speed Across Providers to learn more about DeepSeek R1 pricing and speed across different providers.

Connect to DeepSeek API directly (Recommended)

1. Go to DeepSeek API platform
2. Top up your account to be able to use the API
3. Go to API Keys
4. Click on "Create new API key"
5. Copy the API key
6. Open 16x Prompt and click on the API model settings button in the bottom right corner
7. Go to "3rd Party API / Ollama Configuration" tab
8. Click on "DeepSeek API" button to automatically fill in the API Endpoint for DeepSeek API
9. Enter model ID, for example deepseek-chat (for DeepSeek V3) or deepseek-reasoner (for DeepSeek R1) into the "Model ID" field
10. Click on "Close"

Connect via Hyperbolic

Refer to Hyperbolic Setup Guide on how to setup and use DeepSeek models via Hyperbolic in 16x Prompt.

Connect via Fireworks

Refer to Fireworks Setup Guide on how to setup and use DeepSeek models via Fireworks in 16x Prompt.

Connect via OpenRouter API

Refer to OpenRouter Setup Guide on how to setup and use DeepSeek models via OpenRouter in 16x Prompt.

Connect via DeepInfra

Refer to DeepInfra Setup Guide on how to setup DeepSeek models using DeepInfra in 16x Prompt.

Connect via Ollama

Refer to Ollama Setup Guide on how to setup DeepSeek models using Ollama in 16x Prompt.