AI Video Generator
Advanced options
Powered by CogVideoX (Apache 2.0). Generation takes 30-120 seconds.
What You Can Create
6 Visual Styles
Cinematic, animation, photorealistic, artistic, vintage, and anime. Each style transforms your prompt into a unique visual look.
Flexible Aspect Ratios
16:9 landscape for YouTube, 9:16 portrait for TikTok and Reels, or 1:1 square for social media posts.
2 to 6 Second Clips
Pick 2 to 6 seconds per clip. Free clips run up to 3 seconds; longer clips need purchased tokens or a plan. For a longer film, the Movie tool joins several clips into one MP4.
Negative Prompts
Exclude unwanted elements like blur, text, watermarks, or low quality. Fine-tune your video output with precision.
GPU-Powered
The open-source models run on our own GPU servers, so those prompts are not sent to another company. Premium models are run by the provider that makes them.
MP4 Download, No Watermark
Download the finished clip as an MP4 with no watermark or logo. At busy times a clip waits in a queue before it renders, and the page shows its progress.
How to Use
Describe your video
Write a detailed prompt describing the scene, subjects, actions, and mood. The more specific you are, the better the result.
Choose style and settings
Pick a visual style (cinematic, anime, etc.), set the duration, and choose an aspect ratio that fits your platform.
Refine with negatives
Optionally add things to avoid, like blurry, text, watermark, or low quality. This helps the AI focus on what you actually want.
Generate and download
Click Generate Video and wait for your clip. Preview it in the browser, then download. Not happy? Adjust your prompt and try again.
Token Costs
| Duration | Tokens per video | Best for |
|---|---|---|
| 2 seconds | 30,000 | Quick loops, social media clips |
| 4 seconds | 30,000 | Social media clips, intros, quick animations |
| 6 seconds | 30,000 | B-roll footage, longer scenes, detailed animations |
| 8 seconds | 30,000 | Longest single clip any model allows |
Video Models and What a Clip Costs
A clip is priced per clip rather than per second, because most of the GPU time goes on loading the model and running the diffusion steps, not on the length of the video. The open-source models below run on our own GPUs.
| Model | Best for | Licence | Tokens per video |
|---|---|---|---|
| CogVideoX | The default. Text to video at 480p, on the free tier. | Apache 2.0 | 30,000 |
| Wan 2.2 TI2V-5B | Sharper motion at 720p and 24 fps, from text or from a starting image. Also on the free tier. | Apache 2.0 | 30,000 |
| HunyuanVideo 1.5 | The most detailed of the three, with purchased tokens or a plan. Its licence excludes the EU, the UK and South Korea. | Tencent Hunyuan Community License | 47,000 |
A free account can render one free clip every two days. Premium models from other providers are in the same model menu, and each shows its price before you generate.
More AI Video Tools
Advanced options
Result
❤️ Love Free.ai? Tell your friends!
Sign up to get a referral link and earn 30,000 tokens per friend.
Processing your request...
Generate videos from text and images with AI for free. No watermark. Subtitles + dubbing in 99 languages.
How to Use AI Video Generator
Enter your input
Type text, upload a file, or describe what you want. No account needed.
Click generate
Our AI processes your request in seconds using the best open-source models.
Download & share
Download, copy, or share your result. Free for personal and commercial use.
Use this tool via API
Automate this tool from your own code. OpenAI-compatible REST endpoint, Bearer-token auth, no extra SDK required. Token costs match the web interface.
curl -X POST https://api.free.ai/v1/video/generate/ \
-H "Authorization: Bearer sk-free-..." \
-H "Content-Type: application/json" \
-d '{"prompt": "A cat playing piano", "duration": 4}'AI Video Generator - FAQ
How would you rate this tool?