PRACTICAL GUIDE / 01
How to turn a product photo into an AI video
A practical workflow for preparing a photo, writing a motion prompt, editing a three-shot product video and checking the final result. Includes an adaptable prompt.
Explore the stepsTOPIC GUIDE
AI video generation
New capabilities in Veo, Sora, Runway, Kling and Seedance. Explore official demos, access conditions and the limits of announced features.
Latest news
23 stories
Premiere AI Assistant drops daily limits and moves to Firefly credits
Adobe has removed the daily and weekly free-use caps for AI Assistant in Premiere beta; responses now consume Firefly credits from the user’s plan.
1 min readRead story
iQIYI reconstructs a lost 1938 film with NadouPro
An iQIYI-XMU team rebuilt the 1938 film Visiting Shanghai after Sixty Years in two months using surviving archival images and NadouPro production tools after its prints and script had been lost.
2 min readRead story
Adobe Research presents Project Doki for text-native video authoring
The UIST 2026-honored Project Doki research introduces an experimental video-authoring approach that manages characters, scenes, shots, edits, and audio through a single text document.
2 min readRead story
NVIDIA Cosmos adds real-time interactive world simulation for physical AI
NVIDIA has expanded Cosmos 3 with open interactive world models that respond in real time to camera movement, robot controls and autonomous-driving inputs.
1 min readRead story
Gemini Omni 1.2 Flash adds 15-second video and multi-turn editing
Google’s Gemini Omni 1.2 Flash is generally available with 15-second generation, multi-turn conversational editing, stronger reference consistency and 30% lower pricing.
1 min readRead story
Instagram Edits opens its Mac video tools to AI agents
The new Edits app for Mac lets compatible AI agents assemble rough cuts through MCP, then allows creators to continue editing manually in the same project.
2 min readRead story
Kling 4.0 aims to increase usable footage in long takes
Fountain 0 founder Ash Koosha says early Kling 4.0 tests suggest more consistent physics, spatial relationships and long takes could reduce discarded footage by at least 50%.
2 min readRead story
Synthesia opens AI video creation and editing inside ChatGPT
The public beta of Synthesia MCP lets users create video drafts from ChatGPT, edit Syren videos conversationally and search their existing video library.
2 min readRead story
TBC launches Mischief 1, a Wan 2.2-based AI video model
The Biological Computing Company has made Mischief 1 available through fal and OpenRouter, applying a software optimization derived from neuronal measurements to Wan 2.2.
2 min readRead story
Oracle Media Fabric connects AI content creation with video delivery
Oracle’s planned Media Fabric platform brings generative media tools, enterprise models and large-scale live video delivery into one workflow.
2 min readRead story
Google DeepMind announces Genie 4, a real-time world model with native audio
Genie 4 is a steerable world model that generates visuals, ambient sound, music and interactive dialogue together inside a real-time simulation, initially as a research preview.
2 min readRead story
NVIDIA and TCV invest in Higgsfield’s generative video infrastructure
Higgsfield will use the undisclosed investment for global infrastructure and hiring; its platform combines its own SOUL 2 models with third-party video models in one production workflow.
2 min readRead story
