How I Get Claude To Edit My Videos (Step-By-Step)
Google DeepMind world model used as the subject of an example short; Claude pulled B-roll from the DeepMind site and Matt's own early-access footage.
G4
Genie 4
Google DeepMind world model that turns text or images into interactive, real-time controllable worlds with audio.
Updated EU status checked Last shown
Genie 4 at a glance
- What is Genie 4?
- Genie 4 is a model from Google DeepMind. Google DeepMind world model that turns text or images into interactive, real-time controllable worlds with audio.
- Who makes Genie 4?
- Genie 4 is made by Google DeepMind.
- Does Genie 4 work in the EU?
- Unverified. No explicit evidence found. Access is limited to trusted testers/early access, and no regional availability information was provided. Last verified 14 Nov 2026.
- How much does Genie 4 cost?
- Pricing unknown. Early access for trusted testers; creators say it is coming to Ultra subscribers. No pricing published in the provided material. Licence: Proprietary.
- Where was Genie 4 shown?
- In 3 creator videos by Matt Wolfe, AI Search. Most recently by Matt Wolfe on 4 Nov 2026, at 1:40 in “How I Get Claude To Edit My Videos (Step-By-Step)”.
Sources: the vendor's own pages (linked above) and creator videos at the linked timestamps. EU status is information, not legal advice; see how we check.
About
Genie 4 is the latest generation of Google DeepMind's Genie world-simulation models. It generates interactive environments from text or image prompts, lets users move a character through the scene in real time, and allows new elements to be prompted into the world. This version adds generated audio. DeepMind's site describes it as "a new generation of world simulation".
According to creator reports, it can be paired with Gemini 4 for reasoning. Suggested uses include filmmaking previsualisation, game prototyping and robotics research. One creator who tested early access reported coherence issues, a roughly 2-minute session limit and some crashes.
Access is limited. Creators describe it as restricted to trusted testers or early access, with a rollout to Google AI Ultra subscribers mentioned as upcoming. No public API, weights or technical documentation were found in the provided material.
Key features
- Generates interactive worlds from text or image prompts
- Real-time character control and navigation
- Prompt new elements into a running scene
- Generated audio alongside visuals
- Can be paired with Gemini 4 for reasoning
- Early access with reported ~2-minute session limit
3 moments
As seen on YouTube
Watch the moment: the exact second top creators show it. Videos load only when you press play, via youtube-nocookie.com.
AI finally sleeps, Gemini 4 Carbon, ETH cracked, energy-based models, Anthropic IPO: AI NEWS
World model turning text or images into interactive worlds with real-time control and now audio. Can pair with Gemini 4 for reasoning; limited to trusted testers, coming to Ultra subscribers.
AI News: The Only 6 Stories Worth Knowing This Week
Interactive world model in early access. Users prompt a scene and character, move around, and prompt new elements in, and it now adds sound. The creator tested it and found coherence issues, a 2-minute limit and some crashes, but sees uses in filmmaking previs, game prototyping and robotics.
Keep exploring
Similar AI tools
Closest neighbours by meaning, not just by category.
PG
Project Genie
App · Google DeepMind
Google DeepMind app for generating interactive, explorable worlds from prompts, including places grounded in Street View.
Unverified3D & Worlds
Release date n/a
GW
GWM Worlds 2
Model · Runway
Runway's real-time interactive world model generating 720p/24fps video with audio, steered by text actions and camera input.
Unverified3D & Worlds
3 Sept 2026
LW
Lingbot World 2
Model · Robbyant
Real-time interactive world model from Robbyant that generates controllable, long-horizon environments.
Unverified3D & Worlds
Release date n/a
EV
Evoke
Model
14B autoregressive world model that generates interactive, camera-controlled video worlds in 3 denoising steps.
Unverified3D & Worlds
Release date n/aOpen source
AW
Alaya World
Model
Interactive autoregressive world model generating minute-long playable video with real-time camera control and prompt events.
Unverified3D & Worlds
Release date n/a
WO
Wonder
Model · Adobe
Adobe Research video world model that builds navigable interactive worlds from an image or video at 16 FPS.
Unverified3D & Worlds
Jul 2026