From d69452aef64d7586d1ef8f6cb86f5131f6d37a89 Mon Sep 17 00:00:00 2001
From: =?UTF-8?q?=E6=B3=A0=E9=9F=B3?= <im@ling.plus>
Date: Sat, 31 May 2025 15:16:20 +0800
Subject: [PATCH] fix: update gemini-2.5-flash (#346)

* Update Gemini Model
---
 README.md                                | 2 +-
 src/autoslice/mllm_sdk/gemini_new_sdk.py | 4 ++--
 src/autoslice/mllm_sdk/gemini_old_sdk.py | 4 ++--
 3 files changed, 5 insertions(+), 5 deletions(-)

diff --git a/README.md b/README.md
index a859ef8..70cd706 100644
--- a/README.md
+++ b/README.md
@@ -80,7 +80,7 @@
 - **( :tada: NEW)自动渲染字幕**：采用 OpenAI 的开源模型 `whisper`，自动识别视频内语音并转换为字幕渲染至视频中。
 - **( :tada: NEW)自动切片上传**：根据弹幕密度计算寻找高能片段并切片，该自动切片工具库已开源 [auto-slice-video](https://github.com/timerring/auto-slice-video)，结合多模态视频理解大模型自动生成有意思的切片标题及内容，并且自动上传，目前已经支持的模型有：
   - `GLM-4V-PLUS`
-  - `Gemini-2.0-flash`
+  - `Gemini-2.5-flash`
   - `Qwen-2.5-72B-Instruct`
   - `SenseNova V6 Pro`
 - **( :tada: NEW)持久化登录/下载/上传视频(支持多p投稿)**：[bilitool](https://github.com/timerring/bilitool) 已经开源，实现持久化登录，下载视频及弹幕(含多p)/上传视频(可分p投稿)，查询投稿状态，查询详细信息等功能，一键pip安装，可以使用命令行 cli 操作，也可以作为api调用。
diff --git a/src/autoslice/mllm_sdk/gemini_new_sdk.py b/src/autoslice/mllm_sdk/gemini_new_sdk.py
index 8c0f392..646063b 100644
--- a/src/autoslice/mllm_sdk/gemini_new_sdk.py
+++ b/src/autoslice/mllm_sdk/gemini_new_sdk.py
@@ -12,7 +12,7 @@ def gemini_generate_title(video_path, artist):
     video_bytes = open(video_path, "rb").read()
 
     response = client.models.generate_content(
-        model="models/gemini-2.0-flash",
+        model="gemini-2.5-flash-preview-05-20",
         contents=types.Content(
             parts=[
                 types.Part(text=SLICE_PROMPT.format(artist=artist)),
@@ -22,7 +22,7 @@ def gemini_generate_title(video_path, artist):
             ]
         ),
     )
-    scan_log.info("Using Gemini-2.0-flash to generate slice title")
+    scan_log.info("Using Gemini-2.5-Flash to generate slice title")
     scan_log.info(f"Prompt: {SLICE_PROMPT.format(artist=artist)}")
     scan_log.info(f"Generated slice title: {response.text}")
     return response.text
diff --git a/src/autoslice/mllm_sdk/gemini_old_sdk.py b/src/autoslice/mllm_sdk/gemini_old_sdk.py
index 1023017..eeab7a5 100644
--- a/src/autoslice/mllm_sdk/gemini_old_sdk.py
+++ b/src/autoslice/mllm_sdk/gemini_old_sdk.py
@@ -26,14 +26,14 @@ def gemini_generate_title(video_path, artist):
     prompt = SLICE_PROMPT.format(artist=artist)
 
     # Set the model to Gemini Flash.
-    model = genai.GenerativeModel(model_name="models/gemini-2.0-flash")
+    model = genai.GenerativeModel(model_name="models/gemini-2.5-flash")
 
     response = model.generate_content(
         [prompt, video_file], request_options={"timeout": 600}
     )
     # delete the video file
     genai.delete_file(video_file.name)
-    scan_log.info("Using Gemini-2.0-flash to generate slice title")
+    scan_log.info("Using Gemini-2.5-Flash to generate slice title")
     scan_log.info(f"Prompt: {SLICE_PROMPT.format(artist=artist)}")
     scan_log.info(f"Generated slice title: {response.text}")
     return response.text