Kapwing text to speech: AI Voice, limits and a free alternative
Kapwing is an online video editor that many creators use for Shorts, memes and explainers, and its AI Voice turns a script into a voiceover layer on the timeline. Every plan has a monthly limit on generated speech, and on the free plan it is only a few minutes.
This guide shows how text to speech works in Kapwing according to its help center, what the limits are, and how to voice long scripts for free on your own computer with SuperClonica, then add them to Kapwing.
Text to speech in Kapwing, step by step
According to Kapwing’s help center, open the AI Voice tab in the left sidebar and enter your text. Choose a voice that matches the language of the text; Business and Enterprise users can also use their own voice clones. Optionally add a persona, an AI avatar narrator, or choose None for audio only, then click Add layer. The layer appears on the timeline when generation finishes.
Limits to know
Kapwing’s help article, last updated in December 2024, lists 3 minutes of text to speech for free users, 80 minutes a month on Pro and 200 on Business, with custom limits on Enterprise. The limits renew on the first of the month, whatever your billing date. It mentions 180 voices in more than 40 languages. Plans change, so check Kapwing’s current pricing before planning a long project.
Voice long scripts offline for free
For long videos, voice the script on your own computer. Install the free SuperClonica engine with python -m pip install superclonica-cpu, save the script as a text file and run superclonica --text-file script.txt --voice narrator.json --lang en --normalize --output voiceover.mp3. It runs offline on an ordinary processor with no character limit and no monthly minutes, and --normalize keeps every file at the same loudness. Then upload the MP3 to Kapwing as an audio layer.
Voices Kapwing does not have
Hear catalog voices free on their pages: paste 30 to 100 characters, with no account, and download the line. There are narrators, announcers, game and cartoon characters, and ten stock voices, M1 to M5 and F1 to F5, come with the engine. Voice cloning in Kapwing is for Business and Enterprise; in SuperClonica you can train your own voice in the studio from a clean 5 to 10 second recording, free on your own NVIDIA graphics card.