7.4 KiB
\LlmApi
所有 URI 都相对于 http://localhost
| 方法 | HTTP 请求 | 描述 |
|---|---|---|
| google_file_upload | POST /api/llm/google/vertex-ai/upload | 上传文件到谷歌存储,用于gemini视觉功能 |
| invoke_gemini_ai_api_llm_google_chat_post | POST /api/llm/google/chat | 调用google推理 |
| invoke_media_analysis_api_llm_google_sync_media_analysis_post | POST /api/llm/google/sync/media/analysis | 【同步适合小文件】gemini多模态分析,支持视频,音频,图片 |
| llm_chat | POST /api/llm/chat | 调用大模型进行推理 |
| llm_supported_models | GET /api/llm/model/list | 获取支持的模型列表 |
| llm_task_id | GET /api/llm/task/status | 查询推理过程结果 |
| submit_media_inference_api_llm_google_async_media_analysis_post | POST /api/llm/google/async/media/analysis | 【异步适合需要长时间的推理过程】gemini多模态分析,支持视频,音频,图片 |
google_file_upload
serde_json::Value google_file_upload(local_file) 上传文件到谷歌存储,用于gemini视觉功能
Args: local_file: 本地文件支持 音频,视频, 图片 Returns: { "status": "状态信息, true 成功, false 失败", "data": "gs://fashion_image_block/gallery_v2/mp4/1f76996f-a13e-450e-8c86-029c398932c4.mp4", 分析媒体时候使用该字段 "msg": "", "extra": 原始信息 }
Parameters
| Name | Type | Description | Required | Notes |
|---|---|---|---|---|
| local_file | std::path::PathBuf | 本地文件 | [required] |
Return type
serde_json::Value
Authorization
No authorization required
HTTP request headers
- Content-Type: multipart/form-data
- Accept: application/json
[Back to top] [Back to API list] [Back to Model list] [Back to README]
invoke_gemini_ai_api_llm_google_chat_post
serde_json::Value invoke_gemini_ai_api_llm_google_chat_post(prompt, timeout) 调用google推理
Parameters
| Name | Type | Description | Required | Notes |
|---|---|---|---|---|
| prompt | String | prompt输入 | [required] | |
| timeout | Option<f64> | 超时时间 | [default to 180] |
Return type
serde_json::Value
Authorization
No authorization required
HTTP request headers
- Content-Type: application/x-www-form-urlencoded
- Accept: application/json
[Back to top] [Back to API list] [Back to Model list] [Back to README]
invoke_media_analysis_api_llm_google_sync_media_analysis_post
serde_json::Value invoke_media_analysis_api_llm_google_sync_media_analysis_post(text_prompt, media_uri) 【同步适合小文件】gemini多模态分析,支持视频,音频,图片
Parameters
| Name | Type | Description | Required | Notes |
|---|---|---|---|---|
| text_prompt | String | 文本提示词 | [required] | |
| media_uri | String | gs:xxxx, 上传到谷歌vertex ai 上的链接 | [required] |
Return type
serde_json::Value
Authorization
No authorization required
HTTP request headers
- Content-Type: application/x-www-form-urlencoded
- Accept: application/json
[Back to top] [Back to API list] [Back to Model list] [Back to README]
llm_chat
serde_json::Value llm_chat(prompt, model_name, temperature, max_tokens, timeout) 调用大模型进行推理
Parameters
| Name | Type | Description | Required | Notes |
|---|---|---|---|---|
| prompt | String | prompt输入 | [required] | |
| model_name | Option<String> | 模型名称, 从/model/list接口获取 | [default to gemini-2.5-flash] | |
| temperature | Option<f64> | 模型的温度值 | [default to 0.7] | |
| max_tokens | Option<i32> | 最大token值 | [default to 4096] | |
| timeout | Option<f64> | 超时时间,推理 | [default to 180] |
Return type
serde_json::Value
Authorization
No authorization required
HTTP request headers
- Content-Type: application/x-www-form-urlencoded
- Accept: application/json
[Back to top] [Back to API list] [Back to Model list] [Back to README]
llm_supported_models
serde_json::Value llm_supported_models() 获取支持的模型列表
Parameters
This endpoint does not need any parameter.
Return type
serde_json::Value
Authorization
No authorization required
HTTP request headers
- Content-Type: Not defined
- Accept: application/json
[Back to top] [Back to API list] [Back to Model list] [Back to README]
llm_task_id
serde_json::Value llm_task_id(task_id) 查询推理过程结果
Args: task_id: 任务id Returns: {'status': '值为字符串的时候标识任务运行的状态,为布尔值时,标识任务运行完毕！', 'data': '', 'msg': ''}
Parameters
| Name | Type | Description | Required | Notes |
|---|---|---|---|---|
| task_id | String | 任务id | [required] |
Return type
serde_json::Value
Authorization
No authorization required
HTTP request headers
- Content-Type: Not defined
- Accept: application/json
[Back to top] [Back to API list] [Back to Model list] [Back to README]
submit_media_inference_api_llm_google_async_media_analysis_post
serde_json::Value submit_media_inference_api_llm_google_async_media_analysis_post(text_prompt, media_uri) 【异步适合需要长时间的推理过程】gemini多模态分析,支持视频,音频,图片
Parameters
| Name | Type | Description | Required | Notes |
|---|---|---|---|---|
| text_prompt | String | 文本提示词 | [required] | |
| media_uri | String | gs:xxxx, 上传到谷歌vertex ai 上的链接 | [required] |
Return type
serde_json::Value
Authorization
No authorization required
HTTP request headers
- Content-Type: application/x-www-form-urlencoded
- Accept: application/json
[Back to top] [Back to API list] [Back to Model list] [Back to README]