Groq通过其GroqCloud平台提供AI推理服务，该平台基于LPU（语言处理单元）架构设计，专注于速度和可扩展的性价比。主要功能包括AI模型的高性能推理，强调低延迟和成本效益。该平台支持开发者和企业，提供API访问以构建AI应用。显著优势包括快速的推理速度和有竞争力的定价，如客户案例和演示所示。典型用例涉及各行业的AI驱动应用，例如McLaren F1 Team、Dropbox、Vercel等合作伙伴的展示。服务包括免费API密钥、文档和开发者社区支持。
Groq通过其LPU架构和GroqCloud平台提供快速且低成本的AI推理服务。
api.groq.com完成站点认证，解锁以下权益：
按 3 个模型行对比最新 audit、最新测速、吞吐、延迟与按 token 计费价格。
| 模型 | 输入 ($/M) | 输出 ($/M) | 检测 | 速度 | 延迟 |
|---|---|---|---|---|---|
meta-llama/llama-4-scout-17b-16e-instruct | - | - | — | 447.8 t/s | 0.20 s |
free:Qwen3-30B-A3B | - | - | — | 19.7 t/s | 25.30 s |
| - | - | — | 53.0 t/s | 8.89 s |
当前显示 3 / 3 个模型行
| 时间 | 模型 | 速度 | 延迟 |
|---|---|---|---|
| Apr 25, 11:32 PM | meta-llama/llama-4-scout-17b-16e-instruct | 447.76 tok/s | 0.20s |
| Dec 23, 03:21 PM | free:Qwen3-30B-A3B | 23.23 tok/s | 18.61s |
| Dec 20, 02:59 AM | free:Qwen3-30B-A3B | 24.49 tok/s | 13.14s |
| Dec 18, 08:29 AM | free:Qwen3-30B-A3B | 30.07 tok/s | 9.45s |
| Dec 18, 04:35 AM | free:Qwen3-30B-A3B | 22.38 tok/s | 27.29s |
| Dec 17, 09:28 AM | free:Qwen3-30B-A3B | 24.97 tok/s | 15.03s |
| Dec 16, 12:22 PM | free:Qwen3-30B-A3B | 21.75 tok/s | 4.61s |
| Dec 16, 07:52 AM | free:Qwen3-30B-A3B | 19.47 tok/s | 23.58s |
| Dec 14, 04:14 PM | free:Qwen3-30B-A3B | 29.29 tok/s | 5.65s |
| Dec 13, 03:57 PM | free:Qwen3-30B-A3B | 21.04 tok/s | 1.45s |
排名基于社区提交的测试数据与定期健康探测，仅供参考，非官方数据。
用 89 个 LMSpeed 信号，把 Groq 的 6 个相近 API 替代服务商放在一起比较：共享模型覆盖、价格、实测速度、可用性和免费模型。
| 服务商 | 对比理由 | 模型数 | 免费项 | 均价 | 速度 | 30 天可用性 |
|---|---|---|---|---|---|---|
| Groq groq Groq provides fast and low-cost AI inference through its LPU architecture and GroqCloud platform. | 当前服务商基线 | 5 | 0 | N/A | 20 tok/s | 99.3% |
| OpenRouter openrouter A unified API interface providing access to over 300 models from 60+ providers, including OpenAI, Anthropic, and Google. |
| 221 | 0 | N/A | 86 tok/s | 99.7% |
nvidia-nim NVIDIA NIM provides optimized AI model inference APIs for LLMs, vision, and embedding models through NVIDIA cloud infrastructure. |
| 53 | 0 | N/A | 66 tok/s | 99.5% |
qiniu-2 七牛云提供AI大模型推理服务，包括多种模型调用、智能问答助手、代码助手等企业级AI解决方案。 |
| 40 | 0 | N/A | 42 tok/s | 99.4% |
api-model-yungnet-cn Yungnet Model Open Platform is a China-based unified LLM API gateway with an OpenAI-compatible relay at api.model.yungnet.cn, focusing on flagship Chinese models. |
| 16 | 0 | N/A | 49 tok/s | 99.5% |
huawei-modelarts Huawei ModelArts MaaS (Model as a Service) platform offering Pangu series models and third-party LLMs via OpenAI-compatible API. |
| 10 | 0 | N/A | 30 tok/s | 99.6% |
api-kriora-com Provides OpenAI-compatible APIs and managed GPU instances for deploying and scaling open-source AI models. |
| 9 | 0 | N/A | 565 tok/s | 99.7% |
