Report Bug / Feature Request
AI Video Generator AI Video Editor AI Video Enhancer AI Video Dubbing Studio AI Video Effects — Pikaffects-style AI Video Upscaler AI Shorts Generator AI Reels Generator More →

AI Video Translator

Model:
Repost your YouTube in Spanish, Hindi, French — with your face actually saying the words. Upload a video, pick a target language + voice, and we transcribe, translate, re-voice, and mix the new audio back into the original clip. Toggle lip-sync to also re-render the mouth.

Drag & drop your video here, or click to browse

MP4, MOV, WebM up to 100MB · clips under 5 minutes work fastest

— · — · —
~1,500 tokens per minute of video (audio path) · ~3,000/min with lip-sync.
Extracting audio from video...
Translated video ready
Show transcript + translation
Original (transcribed) 

                    

                    

                        Translated
                        

                    

                

            

            
            

                Layisha phezulu MP4
                
            

            

        

    







    
Where the Video Translator pays off

    

        
Repost YouTube globally
One English upload → Spanish, Hindi, French, Portuguese, Indonesian re-uploads with the speaker's voice in each language. 4× the audience for one production.

        
Localize courses
Online educators dub a 50-video course into 10 languages overnight. The lip-sync option makes the talking-head segments feel native, not subtitled.

        
Localize ads
Run the same product testimonial across 10 country-specific ad accounts. Same actor, same shot, native audio per market — without a reshoot.

    


    
How the Video Translator works

    

        
Step 1
Extract audio
Server-side ffmpeg pulls a clean MP3 track from your upload. Fast — runs on the API VPS.

        
Step 2
Transcribe
Whisper turns the audio into text with auto language detection. 99-language support out of the box.

        
Step 3
Translate + re-voice
MadLAD translates into the target language; Kokoro speaks it back in the voice you chose from 174 options.

        
Step 4
Mix back + (optional) lip-sync
ffmpeg replaces the original audio with the new track. If lip-sync is on, Sync Lipsync v2 re-renders the mouth to match.

    


    
Tips for the cleanest translated video

    
    
        
  • Single forward-facing speaker. Multi-speaker scenes confuse the transcript and the lip-sync pass both.
    • 
        
  • Clear audio at -6 dB. Background music or noise reduces transcription accuracy.
    • 
        
  • Under 5-minute clips render in 2-5 minutes (audio path) or 5-15 minutes (with lip-sync).
    • 
        
  • Spanish, French, German, Portuguese, Italian have the most natural Kokoro voice options. Hindi, Arabic, Japanese, Korean, Chinese also work.
    • 
    






                        
                        

                        
                        
                        
                        
                        

                            Advanced options
                            

                                

                                    
                                

                                

                                    
                                

                            

                        

                    

                    
                


                
                

                    

                        

                            

                                
Imiphumela

                                

                                    
                                    
                                    
                                

                            

                            

                        

                    

                


                
                

                    

                         Tokens running low. 
                        Get More Tokens
                    

                


                
                

                    

                         Want better results? Premium models (GPT-5, Claude, Gemini) deliver higher quality.
                        View Plans
                    

                


                
                

                    
                    

                        
❤️ Love Free.ai? Tell your friends!

                        

                            
                            
                            
                            
                            
                        

                        
                        
Sign up to get a referral link and earn 25,000 tokens per friend.

                        
                    


                    
                    
                    

                        

                            

                                Ufuna okuningi?
                                Sign up free for 10,000 tokens
                            

                            Bhala
                        

                    

                    
                


                
                

                    
                    
Kuqedwa isicelo sakho...

                    

                        

                    

                


                
                
Repost your YouTube in Spanish, Hindi, French — with the speaker's voice in the target language. Whisper transcribes, MadLAD translates, Kokoro re-voices, ffmpeg mixes back. Optional Sync Lipsync v2 re-renders the mouth.

            

        

    







    

        

            

                
Indlela Yokusetshenziswa AI Video Translator

                

                    

                        

                            1
                        

                        
Faka umngenisi wakho

                        
Bhala umbhalo, ulayishe ifayela, noma uchaza ofuna ukukwenza. Akukho akhawunti edingekayo.

                    

                    

                        

                            2
                        

                        
Chofoza ukwakha

                        
I-AI yethu isebenza isicelo sakho emaminithini usebenzisa amamodeli avulekile angcono kakhulu.

                    

                    

                        

                            3
                        

                        
Layisha phezulu futhi uhlukanise

                        
Layisha phezulu, kopela, noma uhlukanise imiphumela yakho. Imahhala ukusetshenziswa komuntu siqu nokwebhizinisi.

                    

                

            

        

    








    

        

            

                
Use this tool via API

                
Automate this tool from your own code. OpenAI-compatible REST endpoint, Bearer-token auth, no extra SDK required. Token costs match the web interface.

                

                    API Documentation
                    Get API Key
                

            

            

                
curl -X POST https://api.free.ai/v1/video/generate/ \
  -H "Authorization: Bearer sk-free-..." \
  -H "Content-Type: application/json" \
  -d '{"prompt": "A cat playing piano", "duration": 4}'

            

        

    










    

        
Amathuluzi ahlobene ne-AI emahhala

        

            
            

                
                    AI Video Generator
                
            

            
            

                
                    AI Video Editor
                
            

            
            

                
                    AI Video Enhancer
                
            

            
            

                
                    AI Video Dubbing Studio
                
            

            
            

                
                    AI Video Effects — Pikaffects-style
                
            

            
            

                
                    AI Video Upscaler
                
            

            
            

                
                    AI Shorts Generator
                
            

            
            

                
                    AI Reels Generator
                
            

            
        

    









    

        

            

                
AI Video Translator — FAQ

                

                    
                    

                        

                            
                        

                        

                            
Upload any video (a YouTube clip, a course lecture, a product testimonial) and we translate it into the target language with the speaker's voice re-synthesized in that language. Optionally, lip-sync re-renders the mouth so it actually matches the new audio. The pitch: repost your YouTube in Spanish, Hindi, French — with your face actually saying the words.

                        

                    

                    
                    

                        

                            
                        

                        

                            
Five steps run client-side in sequence: (1) ffmpeg extracts a clean MP3 from the video, (2) Whisper transcribes it (auto-detects 99 languages), (3) MadLAD translates the text, (4) Kokoro speaks the translation in the voice you picked, (5) ffmpeg mixes the new audio back into the original video, replacing the original track. With lip-sync on, a sixth step routes the result through Sync Lipsync v2 to re-render the mouth.

                        

                    

                    
                    

                        

                            
                        

                        

                            
Video Dubbing always runs the full lip-sync pipeline — it's the premium, identity-preserving option. Video Translator makes lip-sync opt-in: skip it for a 3× faster, ~50% cheaper run that's perfect for voice-over-style content where mouth-perfect sync isn't critical (podcasts, screen recordings, cooking videos, anything where the speaker isn't in close-up the whole time).

                        

                    

                    
                    

                        

                            
                        

                        

                            
The audio-only path runs almost entirely on self-hosted models (ffmpeg + Whisper + MadLAD + Kokoro), so it fits inside your daily token allowance for short clips. Lip-sync uses the premium Sync Lipsync v2 model and requires purchased tokens. Estimate: ~1,500 tokens per minute of video for the audio path, ~3,000/min with lip-sync.

                        

                    

                    
                    

                        

                            
                        

                        

                            
The dropdowns expose 20 high-demand languages (Spanish, French, German, Italian, Portuguese, Dutch, Polish, Russian, Turkish, Arabic, Hebrew, Hindi, Chinese, Japanese, Korean, Vietnamese, Indonesian, Thai, Swedish, English). MadLAD technically supports 450+; ping us if you need others.

                        

                    

                    
                    

                        

                            
                        

                        

                            
Yes — the voice picker pulls all 174 Kokoro voices across 37 languages. Each voice is tagged with a sample so you can audition before committing. Match the voice gender to the speaker for the most natural result.

                        

                    

                    
                    

                        

                            
                        

                        

                            
No — Kokoro picks from 174 stock voices, not a clone of the original speaker. For identity-preserving voice cloning, run our /voice/clone/ tool first to capture the speaker's voice, then use a custom workflow. Voice cloning + auto translation in one pipeline is on the roadmap.

                        

                    

                    
                    

                        

                            
                        

                        

                            
Clips up to 100MB process well. Under 5 minutes renders in 2-5 minutes for the audio path, 5-15 minutes with lip-sync. For longer videos, split into scenes, translate each, and recombine in /video/editor/.

                        

                    

                    
                    

                        

                            
                        

                        

                            
MP4, MOV, WebM, MKV up to 100MB. Single forward-facing speaker gives the cleanest transcript and (if you enable it) the best lip-sync. Background music or multiple speakers reduce transcription accuracy.

                        

                    

                    
                    

                        

                            
                        

                        

                            
Not by this tool — it replaces the audio, not the visuals. To burn translated subtitles into the video, run our /video/caption/ tool on the result. To export an SRT/VTT subtitle file, use /transcribe/ on the original video.

                        

                    

                    
                    

                        

                            
                        

                        

                            
No. Uploads are deleted within minutes of the render. The output sits on our CDN for 24 hours (7 days for paid users) at the share link, then is removed.

                        

                    

                    
                    

                        

                            
                        

                        

                            
Not as a single endpoint — chain the existing live ones: POST /v1/video/to-audio/ → /v1/stt/ → /v1/translate/ → /v1/tts/ → /v1/video/add-audio/, then optionally /v1/video/lip-sync/. Full curl recipes for each at /api/. The frontend orchestrates this exact chain.

                        

                    

                    
                

            

        

    









    

        
Sign up free for 10,000 tokens

        Dala i-akhawunti ekhululekile
        
Akukho khadi le-credit elidingekayo

    









    
How would you rate this tool?

    

        
        
        
        
        
        
        
        
        
        
        
    

    

        
    


    
    








    


    
    



    



    


    

        
Like this tool? Share it!