Understands the video
Uses the title, channel, description, and nearby dialogue to resolve names, terms, and intent.
Public beta · Bring your own API key
YouTube subtitles that
read like someone meant them.
VidLingo uses video context to reshape awkward captions into natural, readable translations—without taking control away from you.
Supports the following AI models
GeminiDeepSeekOpenAIClaudeQwenXiaomi MiMo
RAW CAPTIONSay you want to let your partner know what time you'll get home every time you leave work.
中文假设你想在每次下班时告诉伴侣你什么时候到家。日本語仕事を終えるたびに、何時に帰宅するかをパートナーに知らせたいとします。한국어퇴근할 때마다 파트너에게 몇 시에 집에 도착할지 알려주고 싶다고 가정해 봅시다.ไทยสมมติว่าคุณต้องการบอกคนรักว่าคุณจะถึงบ้านกี่โมงทุกครั้งที่เลิกงาน
Built for watching, not word swapping
Better translation is only half the job.
VidLingo also cleans the caption structure, preserves timing, and remembers the result so the subtitle experience stays smooth from first click to repeat viewing.
Rebuilds broken captions
Turns overlapping auto-captions and fragments into natural subtitle units before translation.
Keeps timing dependable
Derives final timing from cleaned source captions instead of asking the AI to invent timestamps.
Remembers your work
Email sign-in keeps API settings and subtitle history available across your VidLingo sessions.
Three steps
From a YouTube tab to better subtitles.
- 01
Connect your model
Sign in, choose a supported AI provider, paste your API key, and select a model.
- 02
Open a video
VidLingo detects available captions and prefers a manually uploaded track when one exists.
- 03
Generate and watch
Start translation with one click. Completed subtitles load over the YouTube player and stay in your history.
Your key. Your model. Your bill.
VidLingo provides the subtitle workflow—not a hidden AI markup.
The public beta only translates with an API key you provide. Your provider charges your account directly according to its own pricing.
- ✓Keys are encrypted before they are stored.
- ✓Your saved key is only used to call the provider and model you selected.
- ✓VidLingo never promises bundled or free model usage during beta.
Supports the following AI models
GeminiDeepSeekOpenAIClaudeQwenXiaomi MiMo
Questions, answered plainly
Before you install
Do I need my own AI API key?
Yes. During the public beta, VidLingo only supports personal API keys. You choose the provider and model, and your provider bills you directly.
Which AI providers are supported?
Gemini, DeepSeek, OpenAI, Claude, Qwen, and Xiaomi MiMo are currently supported.
Where is my API key stored?
After you sign in and save it, the key is encrypted on the VidLingo backend. It is not displayed again in the extension and is used to run translations with your selected provider.
Does VidLingo work on every YouTube video?
The beta requires an available YouTube caption track. Videos without captions and audio transcription are not supported yet.
What languages can I translate into?
VidLingo currently offers 16 target languages in the extension, including English, Chinese, Japanese, Korean, Spanish, French, and German.
Public beta
Make the next video easier to follow.
Install VidLingo, connect your preferred AI provider, and give one YouTube video a better subtitle track.
Add to Chrome