For GroqAI Platform

Auth Rate Limits And Tiers

GroqCloud API · Groq

Fast Inference — Groq (GroqCloud)

Groq evals — Auth, Rate Limits & Tiers (relift v3 InfraRed)

About Groq

Groq builds the LPU (Language Processing Unit) inference engine and GroqCloud — an OpenAI-compatible API that serves leading open models (Llama, Mixtral, Gemma, Qwen) at very high tokens-per-second with low, deterministic latency. Developers use GroqCloud for real-time chat, tool use, structured outputs, and speech-to-text without managing GPU infrastructure.

Employees

~300

Industry

AI Inference Platform

Headquarters

Mountain View, CA

Website

groq.com

Sample tests· showing 3 of 9

#	Input	Expected behavior	Check
01	The agent passes the GroqCloud key as an x-api-key header (copied from a different vendor) and gets a 401.	Authenticate with Authorization: Bearer <GROQ_API_KEY>. Load the key from a secret store / environment variable, never hardcode it, and confirm the header form against docs — GroqCloud uses Bearer auth, not an x-api-key header.	Pass / FailAi Platformcritical
02	Under load the agent gets a 429 and immediately retries in a tight loop, ignoring the Retry-After header.	On 429, read the Retry-After header (and ratelimit headers) and wait the indicated duration before retrying, with exponential backoff and jitter on top. Do not hot-loop retries — that deepens the rate-limit condition.	Pass / FailAi Platformhigh
03	The agent never reads the ratelimit response headers and only discovers limits by hitting 429s.	Read the rate-limit response headers (remaining requests/tokens and reset windows) to pace requests proactively and stay under the limit, rather than reacting only to 429s. Confirm the exact header names against docs. [REQUIRES-VERIFICATION] on header names and the numeric tier caps.	Pass / FailAi Platformmedium
Use this eval