VideoPoet by Google: Generate and edit video from text prompts
What is VideoPoet by Google?
VideoPoet is a research project from Google that treats video generation as a language modeling problem. A MAGVIT V2 video tokenizer and a SoundStream audio tokenizer convert images, video and audio into discrete codes in a shared vocabulary, and an autoregressive language model predicts the next video or audio token. Training mixes several objectives, including text-to-video, text-to-image, image-to-video, frame continuation, inpainting, outpainting, stylization and video-to-audio. The site presents sample outputs across those tasks: variable-length high-motion clips from text prompts, audio generated to match an input video, longer videos built by predicting one second at a time, camera-motion control through prompt wording, and interactive editing where a user picks from candidate continuations. By default the model outputs 2-second videos. The page lists a paper, a research blog and additional results pages, but no pricing page or sign-up flow.
Focus areas: Text to Video.
Who is VideoPoet by Google for?
- Machine learning researchers studying generative video models
- Research teams exploring multimodal language model training
- Developers evaluating video generation approaches from published research
Is VideoPoet by Google worth it?
A well-documented research model for multimodal video generation, but not a purchasable product
Where VideoPoet by Google is strong
- Covers many video tasks in one model, including text-to-video, image-to-video, stylization, inpainting and video-to-audio
- Longer videos can be produced by repeatedly predicting one second from a one-second input clip
- Camera motion and style can be steered through wording in the text prompt
Where it falls short
- No pricing, sign-up or public product access is published on the site
- Default output is 2-second clips, and the page does not state commercial usage terms
How do you get started with VideoPoet by Google?
- 01Read the linked paper and research blog for the modeling method
- 02Browse the text-to-video, image-to-video, video editing, stylization and inpainting results pages
- 03Watch the sample clips, including the short movie built from generated segments, to judge output quality
VideoPoet by Google: frequently asked questions
What can VideoPoet actually do?
The page describes text-to-video, text-to-image, image-to-video, video frame continuation, inpainting, outpainting, stylization and video-to-audio, and says these tasks can be combined for zero-shot capabilities such as text-to-audio.
How long are the videos it generates?
By default VideoPoet outputs 2-second videos. The page says longer videos are possible by predicting one second of output from a one-second input clip and repeating that process.
How much does VideoPoet cost?
No pricing is published. The site links to a paper, a research blog and results pages, and the pricing URL returns a 404 error.
Can I use it for commercial work?
The page does not state commercial terms, licensing or availability. It presents VideoPoet as a research method with sample results rather than a product you can sign up for.
VideoPoet by Google alternatives
The closest tools in our directory, matched on category, focus areas and audience.
| Tool | Best for | Free plan | Starting price | Focus areas |
|---|---|---|---|---|
| VideoPoet by Google | Researchers exploring language-model-based video generation and editing | Unknown | — | Text to Video |
| Emu Video | Exploring text-to-video generation research | Unknown | — | Text to Video |
| Make-A-Video | Research and concept exploration of text-to-video generation | Unknown | — | Image to Video, Text to Video |
| Mochi 1 AI | Text-to-video generation from written prompts | Unknown | — | Text to Video |
| Flow Video AI | Creators and marketers making short cinematic clips from prompts | No | $9.99/mo | Text to Video |
| Veo 3 Flow | Short AI-generated clips for social, ads and concept work | No | $49.99/mo | Text to Video |
| AIVideoGenerator.me | Quick text-to-video clips from written prompts | Unknown | — | Text to Video |
| Reemix | Making and remixing short videos from text prompts | Unknown | — | Text to Video |
| Text to Video AI | Turning short written scene ideas into video clips | Unknown | — | Text to Video |
| NeverEnds | Anyone exploring text-to-video conversion, pending verification | Unknown | — | Text to Video |
| Sand.ai | Exploring AI models for generating and extending video | Unknown | — | Video Generation |

