Google Gemma 4 31B Instruct 是 Google DeepMind 的 30.7B 级密集多模态模型,支持文本、图像和视频输入,并输出文本。它拥有 256K 令牌上下文窗口,并提供可配置的思考/推理模式。
curl -X POST https://ai.layercloud.org/v1/chat/completions \ -H "Authorization: Bearer YOUR_TOKEN" \ -H "Content-Type: application/json" \ -d '{ "model": "google/gemma-4-31b", "messages": [{"role": "user", "content": "Hello!"}] }'
获取 API 密钥
以纯 Markdown 阅读本页
This page failed to load
A newer version of the site is available. Reload to continue.