No such language.
ElevenLabs accepts a longer clip and lets you explicitly set the language to translate from — a long list, with Bulgarian included. HeyGen accepts a slightly shorter clip, but lets you specify the number of speakers and translate only the audio without touching the picture. For a recording with several people — HeyGen. For long material with a clear source language — ElevenLabs.
The price in credits is per SECOND of video and is shown on the “Generate” button before every run. The “Five seconds cost” row below is the same price multiplied by five.
| What | HeyGen | ElevenLabs |
|---|---|---|
| Model variants on our platform | 1 | 1 |
| What it needs as input | video, language to translate into | video, language to translate into |
| Price in credits per second | 100 | 1205 |
| Five seconds cost | 500 credits | 6025 credits |
| Longest clip at once | 480 sec | 600 sec |
| Sound from the model itself | ✗ | ✗ |
| Negative description (“without…”) | ✗ | ✗ |
| Brand | Variant | Credits per second |
|---|---|---|
| HeyGen | HeyGen Dubbing (translate) | 100 |
| ElevenLabs | ElevenLabs Dubbing | 1205 |
HeyGen can specify how many people speak in the recording. For a conversation, interview or panel, that's the difference between recognized voices and one big jumble.
HeyGen also has a mode in which only the audio is translated, without touching the picture. For cases where the lips must not change, that's exactly the right behavior.
There's also a setting for whether the length of the new audio is adjusted — useful when the translation comes out longer than the original.
ElevenLabs accepts a longer clip at once — see the length row. For a whole lesson or a long interview this means one run instead of splitting.
The source language is set explicitly from a long list instead of relying on detection. For a noisy recording or an accent, this prevents a wrongly detected language.
ElevenLabs also takes part in text voiceover, so one brand covers both dubbing and a voice from scratch.
Where HeyGen loses. It accepts a shorter clip than ElevenLabs and gives no explicit choice of source language — it relies on detection. With an accented recording or background noise, that's a real risk.
Where ElevenLabs loses. It has no setting for the number of speakers. With a multi-person recording the result is less predictable than with HeyGen. There's also no "audio only" mode — if you want the picture left untouched, it can't be set explicitly.
One speaker, a long piece — ElevenLabs. Several speakers or a need to leave the picture untouched — HeyGen.
With both, the price is per second of video. Try a short segment before you run a whole hour — that way you find out how the result sounds before paying for all of it.
HeyGen. HeyGen starts at 100 credits per second, ElevenLabs — at 1205 credits per second. The cheapest option for HeyGen is HeyGen Dubbing (translate), for ElevenLabs — ElevenLabs Dubbing. The exact price is shown on the button before you start the generation.
ElevenLabs. HeyGen goes up to 480 seconds at a time, ElevenLabs — up to 600 seconds. Anything longer is made from several clips stitched together afterwards.
It depends on the mode. HeyGen has an explicit variant where only the audio is translated and the picture is left untouched. If you want lip matching, look at the lip sync models.
See the “Longest file at once” row in the table — both values are read live. ElevenLabs accepts more.
ElevenLabs lets you set the language explicitly, and Bulgarian is in its list. HeyGen relies on automatic detection.
This page compares the two brands with each other. The full technical card for each — all variants, their strengths and weaknesses, typical tasks and real examples from the community — is on its own page.
Last updated:
