Use DeepSeek V4.1 Flash free on InferX, with $0 input and output

InferX is an AI inference platform, and it has just made its DeepSeek V4.1 Flash 1017 Basic endpoint free. Input, output and cached input are all billed at $0, and the endpoint speaks the OpenAI-compatible API, so most clients and plugins can use it once you add the base URL and key.

What you can use

• The free endpoint is DeepSeek V4.1 Flash 1017 Basic, with a 262,144-token context window and no charge for input, output or cached input.
• Capacity on the free endpoint is limited, so requests can be turned away at busy times. Waiting a bit and retrying usually works.
• If you need a longer context, the full DeepSeek V4.1 Flash 1017 endpoint offers about 1M tokens and is currently half price at $0.15 per 1M input tokens and $0.60 per 1M output tokens.

How to use it

1. Sign up for the InferX console and create an inference API key on the API Keys page.
2. Open DeepSeek V4.1 Flash 1017 Basic in the endpoint list, then copy its base URL, model name and your API key into your client.
3. See the InferX pricing page and the quickstart guide for details.