> ## Documentation Index > Fetch the complete documentation index at: https://quwan-b8010aef.mintlify.site/llms.txt > Use this file to discover all available pages before exploring further. # Gemini 多模态对话示例 > 使用Gemini进行图片理解和视频理解的完整示例代码 # Gemini 多模态对话示例 以下示例展示如何使用Gemini的多模态功能，让AI理解和分析图片、视频内容，进行多媒体智能对话。 ## 快速开始 只需要替换 `` 为你的实际API密钥即可运行。 ```bash cURL theme={null} curl -X POST "https://api.tokenops.ai/v1beta/models/gemini-2.5-flash:generateContent" \ -H "Content-Type: application/json" \ -H "Authorization: Bearer " \ -d '{ "contents": [ { "parts": [ { "text": "这张图片里有什么？请详细描述。" }, { "inlineData": { "mimeType": "image/jpeg", "data": "iVBORw0KGgoAAAANSUhEUgAAAAEAAAABCAYAAAAfFcSJAAAADUlEQVR42mNk+M9QDwADhgGAWjR9awAAAABJRU5ErkJggg==" } } ] } ] }' ``` ```python Python theme={null} import requests import base64 import mimetypes import os # 配置API密钥和基础URL API_KEY = "" BASE_URL = "https://api.tokenops.ai/v1beta" def encode_file_to_base64(file_path): """将文件编码为base64格式""" try: with open(file_path, "rb") as file: return base64.b64encode(file.read()).decode('utf-8') except Exception as e: raise Exception(f"文件编码失败: {e}") def get_mime_type(file_path): """获取文件的MIME类型""" mime_type, _ = mimetypes.guess_type(file_path) if mime_type is None: extension = file_path.lower().split('.')[-1] mime_types = { 'jpg': 'image/jpeg', 'jpeg': 'image/jpeg', 'png': 'image/png', 'gif': 'image/gif', 'webp': 'image/webp', 'bmp': 'image/bmp', 'mp4': 'video/mp4', 'mov': 'video/quicktime', 'avi': 'video/x-msvideo', 'webm': 'video/webm' } mime_type = mime_types.get(extension, 'application/octet-stream') return mime_type def gemini_multimodal_chat(text_prompt, file_paths=None): """ 与Gemini进行多模态对话 Args: text_prompt: 文本提示 file_paths: 文件路径列表 (图片、视频) """ url = f"{BASE_URL}/models/gemini-2.5-flash:generateContent" headers = { "Content-Type": "application/json", "Authorization": f"Bearer {API_KEY}" } parts = [{"text": text_prompt}] if file_paths: for file_path in file_paths: if os.path.exists(file_path): file_base64 = encode_file_to_base64(file_path) mime_type = get_mime_type(file_path) parts.append({ "inlineData": { "mimeType": mime_type, "data": file_base64 } }) else: print(f"警告: 文件不存在 - {file_path}") data = { "contents": [{"parts": parts}] } try: response = requests.post(url, headers=headers, json=data) if response.status_code == 200: result = response.json() if 'candidates' in result and len(result['candidates']) > 0: content = result['candidates'][0]['content'] if 'parts' in content and len(content['parts']) > 0: return content['parts'][0]['text'] return f"错误: {response.status_code} - {response.text}" except Exception as e: return f"请求失败: {e}" def analyze_image(image_path): """使用Gemini分析图片内容""" prompt = """请详细分析这张图片，包括： 1. 主要对象和场景 2. 色彩搭配和构图 3. 图片传达的情感 请用中文详细回答。""" return gemini_multimodal_chat(prompt, [image_path]) def analyze_video(video_path): """分析视频内容""" prompt = """请分析这个视频的内容，包括： 1. 视频主题和内容概述 2. 主要场景和画面 3. 人物动作和情节发展 请详细分析并给出专业建议。""" return gemini_multimodal_chat(prompt, [video_path]) # 使用示例 if __name__ == "__main__": print("=== Gemini多模态对话示例 ===\n") sample_image = "sample_photo.jpg" sample_video = "sample_video.mp4" if os.path.exists(sample_image): print("图片内容分析:") print(analyze_image(sample_image)) if os.path.exists(sample_video): print("\n视频内容分析:") print(analyze_video(sample_video)) ``` ## 使用File API上传大视频文件 对于超过20MB的大视频文件，推荐先使用File API上传，然后通过文件URI进行视频理解分析。 ```bash cURL theme={null} # 第一步：上传视频文件 curl --request POST \ --url https://api.tokenops.ai/v1/files/upload \ --header 'Authorization: Bearer ' \ --header 'Content-Type: video/mp4' \ --header 'filename: my_large_video.mp4' \ --data-binary '@/path/to/your/large_video.mp4' # 响应示例: # { # "uri": "tokenops://bucket.example.com/file_api/20241016/my_large_video.mp4", # "mime_type": "video/mp4" # } # 第二步：使用文件URI进行视频理解 curl -X POST "https://api.tokenops.ai/v1beta/models/gemini-2.5-flash:generateContent" \ -H "Content-Type: application/json" \ -H "Authorization: Bearer " \ -d '{ "contents": [ { "parts": [ { "text": "请详细分析这个视频的内容，包括场景、人物、动作和情节发展。" }, { "fileData": { "mimeType": "video/mp4", "fileUri": "tokenops://bucket.example.com/file_api/20241016/my_large_video.mp4" } } ] } ] }' ``` ```python Python theme={null} import requests import os API_KEY = "" BASE_URL = "https://api.tokenops.ai" def upload_video_file(file_path): """上传视频文件到File API""" if not os.path.exists(file_path): raise FileNotFoundError(f"文件不存在: {file_path}") filename = os.path.basename(file_path) mime_types = { '.mp4': 'video/mp4', '.mov': 'video/quicktime', '.avi': 'video/x-msvideo', '.webm': 'video/webm' } file_ext = os.path.splitext(filename)[1].lower() content_type = mime_types.get(file_ext, 'video/mp4') url = f"{BASE_URL}/v1/files/upload" headers = { 'Authorization': f'Bearer {API_KEY}', 'Content-Type': content_type, 'filename': filename } with open(file_path, 'rb') as file: response = requests.post(url, headers=headers, data=file) if response.status_code == 200: result = response.json() return { 'uri': result.get('uri'), 'mime_type': result.get('mime_type') } else: raise Exception(f"上传失败: {response.status_code} - {response.text}") def analyze_video_with_file_api(file_info, prompt="请详细分析这个视频的内容"): """使用上传的文件URI进行视频分析""" url = f"{BASE_URL}/v1beta/models/gemini-2.5-flash:generateContent" headers = { "Content-Type": "application/json", "Authorization": f"Bearer {API_KEY}" } data = { "contents": [{ "parts": [ {"text": prompt}, { "fileData": { "mimeType": file_info['mime_type'], "fileUri": file_info['uri'] } } ] }] } response = requests.post(url, headers=headers, json=data) if response.status_code == 200: result = response.json() if 'candidates' in result and len(result['candidates']) > 0: content = result['candidates'][0]['content'] if 'parts' in content and len(content['parts']) > 0: return content['parts'][0]['text'] return f"分析请求失败: {response.status_code} - {response.text}" # 使用示例 if __name__ == "__main__": video_path = "my_video.mp4" if os.path.exists(video_path): print("正在上传视频文件...") file_info = upload_video_file(video_path) print(f"上传成功，URI: {file_info['uri']}") print("正在分析视频内容...") analysis = analyze_video_with_file_api(file_info, "请描述这个视频的主要内容") print(f"分析结果：{analysis}") ``` ### File API 说明 File API 上传成功后返回包含 `uri` 和 `mime_type` 的JSON格式： ```json theme={null} { "uri": "tokenops://bucket.example.com/file_api/20241016/test.mp4", "mime_type": "video/mp4" } ``` 在视频理解API中使用 `uri` 作为 `fileUri` 字段： ```json theme={null} { "fileData": { "mimeType": "video/mp4", "fileUri": "tokenops://bucket.example.com/file_api/20241016/test.mp4" } } ``` ## API格式说明 ### 基本请求结构 ```json theme={null} { "contents": [ { "parts": [ { "text": "分析这个文件" }, { "inlineData": { "mimeType": "image/jpeg", "data": "base64编码的文件数据" } } ] } ] } ``` ### 多文件输入 ```json theme={null} { "contents": [ { "parts": [ { "text": "比较这些文件" }, { "inlineData": { "mimeType": "image/jpeg", "data": "第一个文件的base64数据" } }, { "inlineData": { "mimeType": "video/mp4", "data": "第二个文件的base64数据" } } ] } ] } ``` This documentation is built and hosted on [Mintlify](https://mintlify.com), a developer documentation platform.