应用实例API¶
API服务端HOST地址为：https://www.autodl.art
鉴权¶
Token获取位置：请登录 www.AutoDL.com 访问控制台 → 账号 → 设置 → 开发者Token
使用Token方式：
headers = {"Authorization": "填写您的token"}
创建应用实例¶
POST /api/v1/adl_dev/dev/instance/pro/create
请求Body示例：
// 默认以按量计费方式创建实例，暂不支持选择其他计费方式创建实例
{
"data_center_list": ["westDC3", "beijingDC2"], // [选填] 默认系统自动选择地区，westDC3：西北区，beijingDC2：北京区
"req_gpu_amount":1, // [必填] GPU数量, 最小值=1，最大值4
"expand_system_disk_by_gb":0, // [必填] 系统盘扩容大小 单位:GB, 取值范围: 0-500
"gpu_spec_uuid":"pro6000-p", // [必填] 算力规格ID，请参考文末附录中不同GPU型号的规格ID
"application_uuid":"J1c7lsvbq5", // [必填] 应用的UUID，打开网页应用详情可以查看
"application_version":"latest", // [必填] 应用的版本, 如果填写latest，则自动使用最新的版本
"instance_name":"API创建的实例" // [选填] 实例备注名
}
返回Body示例：
{
"code": "Success",
"data": "pro-76419909953e", // 创建出的应用实例ID
"msg": "",
"request_id": "983edb8c8c8553db1115d51fb2758ae8"
}
获取应用实例详情¶
GET /api/v1/adl_dev/dev/instance/pro/snapshot
请求Body示例：
{
"instance_uuid":"pro-76576c61fdf1" // [必填] 实例ID
}
返回Body示例：
{
"code": "Success",
"data": {
"region_sign": "bj-B1",
"payg_price": 1970, // 折扣后的按量计费价格
"origin_pay_price": 3030, // 折扣前的按量计费价格
"snapshot_gpu_alias_name": "NVIDIA RTX PRO 6000",
"chip_corp": "nvidia",
"cpu_arch": "x86",
"usage_info": {
"container_id": "autodl-pro-76576c61fdf1",
"valid_at": "2025-12-15T18:35:52.137127391+08:00",
"cpu_usage_percent": 3.34,
"mem_usage_percent": 1.26,
"mem_usage": 270528512,
"mem_limit": 21474836480,
"root_fs_used_size": 54435840,
"root_fs_total_size": 31526391808,
"data_disk_total_size": 0,
"data_disk_used_size": 0,
"storage_fs_usage": "",
"pull_image_progress": 1,
"download_image_progress": 1,
"download_oss_file_progress": 0,
"sys_fs_last_block_size": 0,
"is_new": false,
"valid": false
},
"expand_system_disk_size": 32212254720,
"system_init_disk_size": 32212254720,
"ssh_command": "ssh -p 34222 root@connect.xxx.autodl.com",
"proxy_host": "connect.xxx.autodl.com", // ssh地址
"root_password": "jbeOXgTWUxq+", // ssh密码
"ssh_port": 34222, // ssh 端口
"jupyter_token": "xxx", // jupyterlab token
"jupyter_domain": "a1-765793a1f226.xxx.autodl.com:8443", // jupyterlab的访问地址
"service_6006_domain": "u1-h1tr7dnhvxyvm4uacvq9.xxx.autodl.com:8443", // 6006端口服务对应的访问地址
"service_6006_port_protocol": "http",
"service_6008_domain": "uu1-yufv2v0fcxtvr5lv4j80.xxx.autodl.com:8443", // 6008端口服务对应的访问地址
"service_6008_port_protocol": "http",
"cg_application_info": {
"application_uuid": "vbxoJpZdGD",
"current_version": "v2",
"application_name": "ComfyUI-纯净版",
"logo": "https://codewithgpu-1310972338.cos.ap-beijing.myqcloud.com/1-225165862-7lVX68jiMJQw7QFyqH7f.png"
}
},
"msg": "",
"request_id": "b717410bb575336654fcc2ad79c72675"
}
获取应用实例状态¶
GET /api/v1/adl_dev/dev/instance/pro/status
请求Body示例：
{
"instance_uuid":"pro-76576c61fdf1" // [必填] 实例ID
}
返回Body示例：
{
"code": "Success",
"data": "running", // 实例状态
"msg": "",
"request_id": "2159c05b9b675c731c2334b88d6be8aa"
}
获取应用实例列表¶
POST /api/v1/adl_dev/dev/instance/pro/list
请求Body示例：
{
"page_index":1,
"page_size":1
}
返回Body示例：
{
"code": "Success",
"data": {
"list": [
{
"created_at": "2025-12-15T17:30:54+08:00",
"uuid": "pro-76576c61fdf1",
"machine_id": "4d67438b4f",
"machine_alias": "",
"region_sign": "neimeng-C",
"region_name": "内蒙C区",
"status": "running",
"sub_status": "",
"status_at": "2025-12-15T17:31:05+08:00",
"start_mode": "gpu", // 有卡模式：gpu
"charge_type": "payg", // 计费方式
"req_gpu_amount": 1, // GPU数量
"expired_at": {
"Time": "0001-01-01T00:00:00Z",
"Valid": false
},
"started_at": {
"Time": "2025-12-15T17:31:05+08:00",
"Valid": true
},
"stopped_at": {
"Time": "0001-01-01T00:00:00Z",
"Valid": false
},
"name": "zqAPI创建",
"timed_shutdown_at": {
"Time": "0001-01-01T00:00:00Z",
"Valid": false
},
"gpu_spec_uuid": "pro6000-p", // 算力规格ID
"cg_application_info": {
"application_uuid": "vbxoJpZdGD",
"current_version": "v2",
"application_name": "ComfyUI-纯净版",
"logo": "https://codewithgpu-1310972338.cos.ap-beijing.myqcloud.com/1-225165862-7lVX68jiMJQw7QFyqH7f.png"
}
}
],
"page_index": 1, // 第几页
"page_size": 1, // 每页记录数量
"offset": 0,
"max_page": 22, // 总页数
"result_total": 22, // 总记录数
"page": 1
},
"msg": "",
"request_id": "c53a6dc29de7c65ff2798dac12d8fe78"
}
开机应用实例¶
POST /api/v1/adl_dev/dev/instance/pro/power_on
请求Body示例：
{
"instance_uuid": "pro-759127a8714f", // [必填] 实例ID
"payload":"gpu" // [必填] gpu：有卡开机, 暂不支持API以无卡模式开机
}
返回Body示例：
{
"code": "Success",
"data": null,
"msg": "",
"request_id": "2a8504ac55fece996ace6deeb98e7fe5"
}
关机应用实例¶
POST /api/v1/adl_dev/dev/instance/pro/power_off
请求Body示例：
{
"instance_uuid": "pro-759127a8714f" // [必填] 实例ID
}
返回Body示例：
{
"code": "Success",
"data": null,
"msg": "",
"request_id": "20114f6ef29ab403cc421489bb6e5539"
}
释放应用实例¶
在释放实例前请先关机实例，否则可能无法释放
POST /api/v1/adl_dev/dev/instance/pro/release
请求Body示例：
{
"instance_uuid": "pro-759127a8714f" // [必填] 实例ID
}
返回Body示例：
{
"code": "Success",
"data": null,
"msg": "",
"request_id": "20114f6ef29ab403cc421489bb6e5539"
}
附录¶
GPU型号和算力规格ID对应表
| 前台网页显示GPU型号 | 前台网页显示的规格名称 | API中使用的算力规格ID |
|---|---|---|
| H800-80G | 通用型 | h800 |
| 4090-48G | 通用型 | v-48g |
| PRO6000-96G | 性能型 | pro6000-p |
| vGPU-32G | 性能型 | v-32g-p |
| vGPU-48G-350W | 通用型 | v-48g-350w |
| 5090-32G | 性能型 | 5090-p |