Free API access with per-model limits, e.g. gpt-oss-120b at 30 requests a minute, 1,000 requests and 200K tokens a day
Groq runs fast LLM inference on its own chips and offers it through an API. Its Free Plan gives API access to hosted models like gpt-oss-120b, Qwen and Whisper at no cost, within per-model daily request and token limits.
Verified Oct 251 post on X