Groq is an ultra-fast AI inference platform that leverages custom-designed LPU (Language Processing Unit) hardware to deliver unprecedented inference speeds for open-source LLMs. The platform provides free access to popular models like Llama 2, Mixtral, and Gemma through an OpenAI-compatible API, making it easy for developers to integrate blazing-fast AI capabilities into their applications. Groq's custom hardware enables token generation speeds up to 10x faster than traditional GPUs, with a generous free tier and competitive pay-per-use pricing for production workloads requiring maximum performance.
Review freshness
Mar 13, 2026
Last checked Feb 2, 2026
Next review Apr 3, 2026 · overdue
Confidence
65%
Completeness: partial
Update method
official baseline
Official-source-first profile maintenance
Cadence every 21 days
Sources
5
Baseline: complete
Live verification
Partially verified
Verified required types: 2
Last check Apr 16, 2026
官方网站优先给你结论，再给你命令，不再把 provider 详情页当成旧 CLI 指令堆。
运营模式
Cloud platform access
接入建议
Recommended
先让 CCJK 检测环境，再结合本页的运营模式和接入建议决定是否启用该 provider。
npx ccjkCI、批量机器或标准化部署时使用静默模式，保持和官方 README 一致。
export ANTHROPIC_API_KEY="sk-ant-..." && npx ccjk init --silent完成初始化后再启用环境优化与开发权限 preset，不把 provider 接入和基础环境耦合在一起。
npx ccjk boost && ccjk zc --preset dev先读完本页的官方来源、运营模式和接入结论，再导出配置。对 relay、聚合商或转售通道，建议使用隔离 API key、独立预算和单独环境。
导出环境变量模板
curl -fsSL "https://www.claudehome.cn/api/providers/groq/config?format=env"导出 curl 示例
curl -fsSL "https://www.claudehome.cn/api/providers/groq/config?format=curl"导出 OpenAI 兼容配置
curl -fsSL "https://www.claudehome.cn/api/providers/groq/config?format=newapi"Groq is an ultra-fast AI inference platform that leverages custom-designed LPU (Language Processing Unit) hardware to deliver unprecedented inference speeds for open-source LLMs. The platform provides free access to popular models like Llama 2, Mixtral, and Gemma through an OpenAI-compatible API, making it easy for developers to integrate blazing-fast AI capabilities into their applications. Groq's custom hardware enables token generation speeds up to 10x faster than traditional GPUs, with a generous free tier and competitive pay-per-use pricing for production workloads requiring maximum performance.
Classification
Cloud platform access
Groq operates its own managed inference platform and commercial layer rather than acting as a simple pass-through relay.
Procurement
Platform account with usage billing and enterprise sales options
Integration Pattern
GroqCloud managed inference API
Risk Signal
medium
Good production candidate when low-latency managed inference on GroqCloud matters more than direct control over every open model host. Live verification is currently partial because some required official source types are blocked from this environment: documentation, pricing.
openaiVerified required types
Blocked or broken types
Completeness
complete
Groq provides a complete public baseline across GroqCloud docs, pricing, support, and terms.
Covered types
Missing types