Translate Audio voiceover failed or is stuck
Fix voiceover errors in Translate Audio: ttsFailed, short audio, low balance, incompatible provider and stuck jobs.
Written By Umakhan Magomedov
Last updated 5 days ago
If voiceover generation fails, shows an error toast, or seems stuck, this guide covers the most common causes and fixes in Translate Audio.
Common symptoms
ttsFailed toast: a short error message at the top when you tap play
Audio too short: cloning providers refuse very short source files
Insufficient balance: not enough tokens on your account
Provider incompatible: voice provider does not support the target language
Stuck job: progress spinner runs for a long time without finishing
Voiceover failed (ttsFailed)
Check your token balance on the Profile tab. Top up if needed: Not enough tokens.
Open Settings and verify the voiceover provider supports your target language. See Translate Audio settings.
If using MiniMax or Qwen cloning, confirm the uploaded audio is long enough (MiniMax: 10+ sec, Qwen: 3+ sec).
Tap play again. For HeyGen, wait up to 10 minutes on long texts.
If the error repeats, switch to ElevenLabs for a quick standard voiceover test.
Audio too short for cloning
Voice cloning needs enough speech in the source file:
MiniMax: minimum 10 seconds of clear speech
Qwen: minimum 3 seconds
If the file is shorter, the app auto-switches to ElevenLabs. You can also pick ElevenLabs or OpenAI manually in Settings. For better clone quality on longer files see Voice cloning failed or sounds wrong.
Insufficient balance
Voiceover charges tokens when generation starts. MiniMax may also charge a 150-token clone activation fee the first time a voice is used. If balance is too low, generation fails immediately.
Fix: buy tokens or subscribe to Premium. Details: Not enough tokens and Token pricing for each tool.
Provider incompatible with language
Qwen supports only 10 languages. HeyGen and MiniMax have their own language lists. When you change the target language to one the provider does not support, VocaLingo switches to ElevenLabs automatically and shows a brief notice.
To avoid surprises, check provider language support in Translate Audio settings before generating audio.
Job stuck or still processing
HeyGen: can take ~10 minutes. Leave the app open or reopen the result from History later. Completed jobs resume automatically.
MiniMax / Qwen: background jobs continue on the server. Close and reopen the History entry to refresh status.
If stuck for more than 15 minutes, switch provider to ElevenLabs, tap play again and contact support if charges appeared without audio.
Frequently asked questions
I was charged but hear no audio. What now?
I was charged but hear no audio. What now?
Reopen the result from History. If a background job completed, audio should appear. If balance was deducted but audio is missing after 15 minutes, check Usage History in Profile and contact support with the timestamp.
Why does the player reset after I edit the translation?
Why does the player reset after I edit the translation?
Editing clears outdated audio so you do not hear mismatched speech. Tap play to generate fresh voiceover. This is expected behavior, not an error.
HeyGen finished but I only see an MP4 file. Is that normal?
HeyGen finished but I only see an MP4 file. Is that normal?
Yes. HeyGen v3 returns audio as MP4. The in-app player handles it the same as MP3 from other providers.
Can I retry without paying the 150-token MiniMax fee again?
Can I retry without paying the 150-token MiniMax fee again?
Yes, if you use the same activated voice. The 150-token fee applies once per unique voice. Retries with the same clone cost only 0.15 tokens/sec.
Qwen charged 5 tokens for a very short phrase. Why?
Qwen charged 5 tokens for a very short phrase. Why?
Qwen has a minimum charge of 5 tokens per request, even when the spoken output is only a few seconds long.
Should I contact support?
Should I contact support?
Contact support if errors persist after trying ElevenLabs, your balance is correct, and the source audio is at least 10 seconds with clear speech. Include the approximate time of the failed attempt.
Related articles
Was this helpful?
More in Troubleshooting
I didn't receive the login codeFile format not supported or file too largeVideo processing takes too long or failedNot enough tokensStill need help? Share an idea