Genie 4 builds on capabilities from Google DeepMind's Gemini Omni video model to render realistic scenes, while Gemini models handle background planning to maintain world coherence

Overview Debates Related claims Topics Sources

Inkling 2 Small includes a vision encoder accounting for 292 million parameters, allowing it to process images and text while outputting text, but does not process audio inputs

Gemini Omni 1.2 Flash introduces video duration control, allowing users to generate clips in one-second increments between three and 15 seconds

Gemini Omni 1.2 Flash increases maximum native video clip generation and editing length to 15 seconds, up from 10 seconds in Omni 1.1 Flash

Gemini Omni 1.2 Flash supports conversational editing, enabling turn-by-turn modifications of background, lighting, aspect ratio, or specific details without regenerating the full clip

Google DeepMind released Gemini Omni 1.2 Flash, making the multimodal video generation and editing model generally available across Google Cloud, the Gemini API, Google AI Studio, and related applications

Google said Gemini Omni 1.2 Flash allows users to combine character, product, or style images with motion clips of up to three seconds to maintain visual consistency across camera transitions

Google Vids users can generate 15-second video clips using Gemini Omni 1.2 Flash at no additional cost

Google reduced pricing for Gemini Omni 1.2 Flash by 30 percent compared to Omni 1.1 Flash, setting the rate at $0.07 per second for 720p resolution video

ByteDance's Seedance 2.0 model supports native single-clip generation up to 15 seconds, matching Gemini Omni 1.2 Flash's maximum clip duration

Google DeepMind's Veo 3.1 video model generates single video clips of four, six, or eight seconds, compared to the 15-second capability of Gemini Omni 1.2 Flash

Google DeepMind stated that Gemini Omni 1.2 Flash still experiences performance limitations with fast sports action, rapid script modifications, and complex physics

A full 15-second 720p video clip generated using Gemini Omni 1.2 Flash costs approximately $1.05

ByteDance Ltd. maintains market dominance in chatbots and video generation, while rival Chinese tech giant Alibaba Group Holding Ltd. prepares to launch its next-generation foundation model Qwen4

Former OpenAI researcher Yao Shunyu, who joined Tencent to oversee the Hunyuan series, revamped Tencent's research team and brought in former colleague Tian Yonglong to lead multimodal AI development

Tencent stated that the Hy4 model introduces visual understanding, enabling it to execute tasks across command-line and graphical desktop interfaces

The Liquid AI LFM3 family includes LFM3-1.7B for text agents, LFM3-VL-2B for vision, LFM3-Audio-2B for speech, LFM3-Omni-2B for multimodal input, and the d1-2B and d1-Omni-2B decision models

AI video startup Runway released its second playable world model, GWM Worlds 2, featuring dynamic user interaction and audio generation prior to the release of Genie 4

Google DeepMind tested its Gemini 4 Argon model alongside Genie 4 to simulate an Apptronik Apollo humanoid robot playing checkers without using physical hardware

Google DeepMind integrated Gemini 3.8 Flash into Genie 4 as a guiding agent that evaluates incoming frames, breaks down goals into steps, and keeps simulations on track

Genie 4 is Google DeepMind's first real-time world model capable of generating audio and images together dynamically in response to user inputs