Using DeepSeek With PostQode

PostQode supports accessing models through the DeepSeek API, including deepseek-chat and deepseek-reasoner.

Website: https://platform.deepseek.com/

Getting an API Key

Sign Up/Sign In: Go to the DeepSeek Platform. Create an account or sign in.
Navigate to API Keys: Find your API keys in the API keys section of the platform.
Create a Key: Click "Create new API key". Give your key a descriptive name (e.g., "PostQode").
Copy the Key: Important: Copy the API key immediately. You will not be able to see it again. Store it securely.

PostQode supports the following DeepSeek models:

deepseek-chat (Recommended for Coding) — Powered by the DeepSeek-V3 671B Mixture-of-Experts (MoE) architecture. Optimized for surgical diff generation, multi-file code editing, and tool use.
deepseek-reasoner (Recommended for Planning) — Powered by DeepSeek-R1. Employs large-scale reinforcement learning for long chain-of-thought reasoning, mathematical logic, and complex architectural diagnosis.

Open PostQode Settings: Click the gear icon () in the PostQode panel.
Select Provider: Choose "DeepSeek" from the "API Provider" dropdown.
Enter API Key: Paste your DeepSeek API key (sk-...) into the "DeepSeek API Key" field.
Select Model: Choose deepseek-chat or deepseek-reasoner from the "Model" dropdown.
(Optional) Custom Base URL: By default, PostQode connects to https://api.deepseek.com/v1. You can customize this if routing through a proxy.

Native Context Caching: DeepSeek automatically caches repeated prompt prefixes on server clusters, reducing input token costs by up to 90% for cached context.
Thinking Budgets & Output: With deepseek-reasoner, PostQode streams chain-of-thought traces directly into the Plan Mode interface for full transparency before diff application.
Pricing: DeepSeek provides industry-leading cost efficiency (~$0.14–$0.55 per 1M input tokens). Refer to the DeepSeek Pricing page for current rates.