HeyGen Brings Its Video Model and Voice Cloning to fal
HeyGen Video is now on fal, generating 5–15 second clips at up to 2K with built-in sound from $0.02 a second. HeyGen Voice adds speech in 39 languages with voice cloning.
HeyGen’s video and voice models are now available on fal. HeyGen Video generates 5–15 second clips at up to 2K with dialogue, ambience, and sound effects built in. HeyGen Voice adds text-to-speech in 39 languages with instant voice cloning.
HeyGen is best known for its AI avatar videos. HeyGen Video is its first general-purpose video model, built to generate full scenes rather than talking-head presenters.
HeyGen Video
HeyGen Video is available on fal in three modes: text-to-video, image-to-video, and reference-to-video. Reference mode accepts images, video, and audio that you can point to directly in your prompt.
Every clip includes generated sound, from dialogue to background ambience and effects. Clips run 5 to 15 seconds at 480p, 768p, 1080p, or 2K.
HeyGen pitches the model for short business videos with a lot of motion, such as brand moments, product demos, and training content. HeyGen says its highest-quality setting ranked first on Artificial Analysis’ silent text-to-video leaderboard.
HeyGen Voice
HeyGen Voice turns text into speech using either a library voice or an instant clone made from a reference clip. fal says it supports 39 languages and offers a faster Turbo tier.
Output is a WAV file, which makes it easy to drop into an edit alongside generated or live-action footage.
Pricing on fal
HeyGen Video costs $0.02 per second at 480p and 768p and $0.06 per second at 1080p and 2K. A 10-second clip runs $0.20 at 768p or $0.60 at 2K.
HeyGen Voice costs $0.03 per 1,000 characters of input text.
Competitive Context
HeyGen Video’s pricing is lower than most of the video models on fal. FLUX 3 Max is $0.17 per second at 720p at its regular price, covered in fal Launches FLUX 3 Max, and Seedance 2.5 runs roughly $0.47 per second at 720p on fal.
Native audio is now standard among top models, including Seedance 2.5, FLUX 3, and MiniMax H3. For background on how it works, see What Is Native Audio in AI Video?.
On the voice side, HeyGen Voice competes with ElevenLabs and Google’s text-to-speech models for voiceover and dubbing work.
The Signal in the Noise
For anyone producing a lot of short clips, the low per-second price and 2K output make HeyGen Video easy to test against other models. Its focus on business and product content may make it a better fit for commercial work than narrative filmmaking.
Voice cloning is useful for temp voiceovers and localization, but it should only be used with voices you have permission to clone.
Specs & Pricing
- HeyGen Video modes: Text-to-video, image-to-video, reference-to-video
- Clip length: 5–15 seconds
- Resolution: 480p, 768p, 1080p, 2K
- Audio: Dialogue, ambience, and effects built in
- Video pricing: $0.02/sec (480p, 768p), $0.06/sec (1080p, 2K)
- HeyGen Voice: 39 languages, instant voice cloning, Turbo tier
- Voice pricing: $0.03 per 1,000 characters
Which video model are you reaching for most right now, and would price change that? Share in the comments.
