connections
GET /v1/inference/deployments
The organisation's model deployments and what is actually serving.
身份验证
请将 API 密钥作为 bearer 令牌发送。此端点在规范中未指明具体的权限,因此请为您的密钥赋予所需的最小权限,并通过检查响应来确认,而不是盲目假设。
此端点不需要组织 ID。您的密钥已用于识别其所属的组织,且响应范围也仅限于该组织。
免费试用
将尖括号中的内容替换为您自己的值,并将键占位符替换为您仪表板中的一个键。
curl -X GET https://api.zinndigital.com/v1/inference/deployments \
-H "Authorization: Bearer zdk_live_…"已登录?您仪表板中的 API 控制台会自动填入您真实的组织 ID 和您自己的密钥,并针对实时 API 运行请求,以便您查看实际的响应。 在 API 控制台中打开此端点
详细信息
Model endpoints on the customer's own inference account. ⛔ replicas is what was asked for and ready_replicas is what is serving. A screen showing only the first reports a healthy deployment that is answering nothing. ready_replicas is null when the provider did not report readiness at all — which is not the same as zero. state includes degraded (some replicas up, some not) and deleting (being torn down, and possibly still holding GPUs). Neither is rounded onto a neighbour: a degraded deployment reported as running hides missing capacity, and a deleting one reported as stopped implies it has stopped costing money.
响应
| 名称 | 类型 | 必填 | 内容简介 |
|---|---|---|---|
deployments | InferenceDeployment[] | 是 | — |
此端点可能返回的错误
401 · 429