Upload a video and get it back in another language: same voice, lips re-synced to the new speech. 30 languages, no subscription — pay per minute.
This isn't subtitles or voice-over dubbing: the AI translates the speech, clones the speaker's voice, delivers the translation in that same voice — and then re-syncs the lip movements to the new speech. Viewers see a person who appears to speak the language fluently. Built on HeyGen technology — the world leader in video translation.
The main use case is reaching new audiences: translate your videos into Spanish, Chinese or Arabic for new markets — no reshoots, no voice actors, no studio. It works in any direction: localize courses, ads and interviews while keeping the original speaker's voice and face.
HeyGen direct sells subscriptions from $29/mo on top of roughly $2 per translated minute. NeuralBox has no subscription: upload a video (up to 500 MB) and pay for the minutes — about $2.50 per minute of footage, all-in. Pay by card (Stripe) or crypto; tokens never expire.
Upload a video and pick a language — the AI does the rest.
| Target languages | 30: English, Spanish, Chinese, German, French, Arabic, Hindi, Japanese and more |
| File size | up to 500 MB |
| Voice | cloned automatically from the original |
| Lip sync | yes — articulation adapts to the translation |
| Billing | per minute of footage (rounded up) |
| Technology | HeyGen Video Translate |
| API access | yes — the NeuralBox API |
| Video length | Our price | ≈ USD | HeyGen direct |
|---|---|---|---|
| 1 minute | 75,000 tokens | ≈ $2.50 | $2 + subscription from $29/mo |
| 3 minutes | 225,000 tokens | ≈ $7.50 | $6 + subscription |
| 10 minutes | 750,000 tokens | ≈ $25 | $20 + subscription |
Billing is per minute of footage: 75,000 tokens per minute. Token rate shown for the Basic plan ($5 = 150,000 tokens); larger plans cut token cost by up to 40%. HeyGen direct: a subscription from $29/mo plus ≈ $2 per translated minute.
Subtitles have to be read; dubbing plays a stranger's voice over the original. Here the translation is spoken in the speaker's cloned voice, and the lips on screen move in sync with the new speech — as if the person filmed the video in that language.
Yes, the AI clones the voice from the original audio track: timbre and delivery are preserved. The cleaner the source audio, the more faithful the copy.
30 languages: English, Spanish, French, German, Chinese, Japanese, Korean, Arabic, Portuguese, Hindi, Turkish and more. Translation works in any direction.
About $2.50 per minute of footage at the Basic plan rate (down to about $1.50 on larger plans). No subscription — HeyGen direct requires $29/mo plus roughly $2 per minute.
By card via Stripe or with cryptocurrency. You buy tokens, not a plan — one balance also covers chat, image and audio tools.
No. Tokens stay on your balance until you spend them — no monthly reset.
Depends on the video length — usually a few minutes to half an hour. The finished video appears in your history; you don't have to keep the page open.
Videos with a person speaking on camera and clean audio: vlogs, courses, presentations, interviews. Files up to 500 MB.
The technology is primarily designed for a single speaker; quality may drop on multi-voice recordings — test on a short clip first.
Yes, video translation is available via the NeuralBox API — docs at neuralbox.io/api.
Your voice, your video — in 30 languages. No subscription, pay per minute.
Get started