Tools
Tools I used on real tasks come first. Cards marked "Not tested by this site" summarize other people’s reviews. Free tools are listed too, because sometimes the free one is good enough.
Open-source text-to-speech model from Alibaba's Qwen team. It can clone a voice from a short recording.
Good enough for my video narration once I switched to x-vector mode. Free, but slow on a CPU and it needs a Python setup.
- Free, runs on your own computer
- Works on a laptop with no GPU
- Fluent narration in x-vector mode
- About 3 to 6 seconds of computing per second of speech on a CPU
- Needs Python and some setup
- Sometimes misreads Chinese characters that have several readings
- Sounded less natural than ElevenLabs in a side-by-side listen
AI coding assistant from Anthropic that works on your computer. You describe the task in plain language; it writes and runs the code.
This site and the scripts in the voice-cloning test were built with it. You describe the job in plain language, check the result and ask for changes.
- Turns plain-language requests into working scripts
- Runs the code and fixes its own errors
- Useful beyond coding: documents, video pipelines
- Paid
- You still have to check its work
- Long jobs can use up the usage limit, then you wait for it to reset
Online AI voice generator with a large voice library, including 2,500+ Chinese voices.
Gave the most natural Chinese narration in my side-by-side test, in seconds. The free plan is enough to try it, but it has no commercial license and no voice cloning.
- Stock Mandarin voices sounded the most natural in my test
- Two 35-second takes in under 8 seconds
- Chinese uses few credits: about half an hour a month on the free plan
- Free plan: no commercial use and no voice cloning
- Text and audio are processed in the cloud
- One take may have misread a word
AI avatar video generator. You type a script and a stock avatar or your own digital twin presents it. It also translates videos with lip sync.
The reviewers who make short social videos preferred it for realistic avatars and creative options. Their common warning is the credit system, so work out how many minutes your plan covers before you pay.
- 700+ stock avatars on paid plans
- Digital twin from a short recording, with extra looks
- Lip-sync translation into 175+ languages on paid plans
- Avatar IV with a video avatar uses 31 credits a minute, about 19 minutes on Creator
- Creator cannot buy extra credits
- 4K export only on Pro
AI video platform aimed at companies: avatar presenters, an AI assistant that drafts videos from a prompt or document, dubbing, quizzes and team review.
The reviewers who make training and company videos preferred it for polished presenters, team features and a cheaper entry plan. A creator reviewer found voice tuning slower than in HeyGen.
- Starter covers about 150 minutes a year, plus 10 free minutes a month
- AI assistant drafts a whole video from a prompt, PowerPoint or PDF
- Translation glossary keeps product names fixed
- Brand kits, SCORM export and unlimited minutes are Enterprise only
- 125+ stock avatars on Starter, 240+ only on Enterprise
- Unused monthly credits do not roll over