Question 1

How do I use the Groq API?

Accepted Answer

Create a China OpenRouter API key and point any OpenAI-compatible SDK at https://p.nat6.net/v1. Set the model to gpt-oss-20b. On pay-as-you-go keys, groq/gpt-oss-20b pins every request to Groq; DevPass coding plans do not support provider pinning. You do not need a separate Groq account or SDK.

Question 2

Which Groq models are available?

Accepted Answer

China OpenRouter serves 2 Groq models: GPT OSS 20B, GPT OSS 120B. Each model page lists per-token pricing, context window and capabilities.

Question 3

Does Groq train on API data?

Accepted Answer

No. Groq's published policy states that API traffic is not used for model training.

Question 4

Where is Groq based?

Accepted Answer

Groq is headquartered in United States. China OpenRouter shows each provider's location and data policy so you can route by compliance requirements.

Question 5

What happens if the Groq API is down?

Accepted Answer

When you request a model without a provider prefix, China OpenRouter retries failed requests on other providers that serve the same model, so an outage at Groq does not have to take your app down. Send X-No-Fallback: true with a provider-pinned model when every request must reach Groq.

Groq API

Data & Privacy

Available Models

GPT OSS 20B

GPT OSS 120B

Groq API questions