Groq

1. Overview and quickstart

   Groq's LPU-based inference API and OpenAI-compatible access.

2. Supported models

   Hosted models for text, speech, vision, and reasoning.

3. Text and chat completions

   Chat completion parameters, system prompts, and streaming.

4. Tool use

   Function calling and parallel tool use.

5. Structured outputs

   JSON schema mode for predictable output.

6. Reasoning

   Reasoning models and controlling reasoning effort.

7. Speech to text and text to speech

   Whisper transcription and speech generation.

8. Vision

   Image inputs for multimodal models.

9. Batch processing

   Asynchronous large-scale jobs at lower cost.

10. Prompt caching

    Automatic prefix caching to cut latency and cost.

11. Rate limits

    Requests and tokens per minute and handling 429 responses.

Each section links to the official page. Titles and summaries by Extra Credit; the docs belong to their authors.