feat: update README files to clarify translation capabilities and optimize for Japanese

This commit is contained in:
lml committed 2026-03-21 22:49:45 +08:00
1 parent 31adb05ec8
commit ddd462af7f
2 files changed
+16 -19

No files matched your search

+11 -13
View File
@@ -9,8 +9,9 @@
</p>
# What is Caption Trans?
Import videos, extract subtitles, and use AI to translate them into the target language.
Supports: Google Gemini (tested), OpenAI, DeepSeek, Ollama, and other OpenAI-compatible API services.
Use AI large language models to translate video subtitles. **Especially optimized for Japanese.**
Supports: Google Gemini, OpenAI, DeepSeek, and other OpenAI-compatible API services.
# Download
Supports macOS (Apple Silicon) and Windows.
@@ -18,22 +19,19 @@ Supports macOS (Apple Silicon) and Windows.
Please go to [Releases](https://github.com/cddqssc/Caption-Trans/releases) to download.
## ⚠️ How to open on macOS
1. Double-click the app. It will be blocked by the system because it has not been signed with an Apple developer certificate.
1. Double-click the app. Since it is not currently signed with an Apple developer certificate, it will be blocked by the system.
2. Go to **System Settings** > **Privacy & Security**.
3. Scroll down to the Security section and click **"Open Anyway"** next to the app's block message.
4. Enter your Mac password and click **"Open"**.
*(You only need to do this once.)*
3. Scroll down and find the blocked app notice in the Security section, then click **"Open Anyway"**.
4. After verifying your Mac password, click **"Open"** in the final popup.
*(You only need to do this once. After that, you can open it normally by double-clicking.)*
# App Screenshot
# App Screenshots
<img src="screenshots/screenshot_en.jpg" alt="App Screenshot" width="500">
# Tips
I have currently tested the Gemini API, Qwen, and SiliconFlow, and I recommend using **gemini-2.5-flash-lite**.
# Notes
Transcription is especially optimized for Japanese, and GPU acceleration is supported on both Windows and macOS.
It balances speed and quality, and the price is relatively low. It can also translate sensitive content.
For models above Gemini 3, translating sensitive content may return an empty result, and the price is higher.
For translation, I recommend the **gemini flash lite** model. In practice, it is very fast, delivers solid quality, and is relatively affordable. It can also translate some sensitive content.
# License
[MIT License](LICENSE)
+5 -6
View File
@@ -9,8 +9,9 @@
</p>
# 什么是 Caption Trans?
导入视频,提取视频字幕,使用 AI 翻译字幕为目标语言。
支持:Google Gemini(已测试)、OpenAI、DeepSeek、Ollama 等兼容 OpenAI 的 API 服务。
使用 AI 大模型翻译视频字幕。**特别为日语优化。**
支持:Google Gemini、OpenAI、DeepSeek、等兼容 OpenAI 的 API 服务。
# 下载
支持 macOS(M系列芯片)和 Windows。
@@ -28,11 +29,9 @@
<img src="screenshots/screenshot_zh.jpg" alt="应用截图" width="500">
# 使用心得
目前测试了 gemini api, Qwen, SiliconFlow,推荐使用**gemini-2.5-flash-lite**。
转录为日语特别优化,在 windows和 macOS 均支持 GPU 加速。
它能兼顾速度和质量,价格也比较便宜。并且能翻译敏感内容。
gemini-3 以上的模型,翻译敏感内容会返回空,价格也比较贵。
翻译推荐使用**gemini flash lite**模型,实测翻译速度非常快和质量也很不错,价格也比较便宜。并且能翻译一些敏感内容。
# 开源协议
[MIT License](LICENSE)