Turn a supported web page, script, source text, or image collection into a narrated, captioned video.
Words and Video combines source imagery, approved assets, and AI-generated visuals inside one configurable production workflow.
A place-based civic story: scene plan, narration, timed captions, music, and title cards rendered in one pass.
Kiryat Hadarim — Ruti Almog Garden. A fictional municipality, built from a prepared story and the project's own image set.
A property listing may contain dozens of useful photographs. An article may contain one. A script may begin with no visuals at all.
Words and Video is designed around that reality. The system builds a scene plan, uses source images where they fit, gives priority to approved images supplied by the user, and generates new visuals where they are needed.
Narration, captions, pacing, music, and final assembly remain part of the same workflow.
Provide a supported article or property-listing URL, paste a script or source text, upload your own images, or combine them. Different video formats use different inputs.
Supported articles and property listings.
For explainers, documentaries, briefings, and other prepared formats.
Uploaded photographs, approved library assets, interface captures, or other project visuals.
Some workflows can extract text, information, and images from supported pages. Others begin with a prepared script and visual plan. If a page cannot be extracted reliably, a project can still begin with supplied text and images.
For supported article and real-estate listing pages, Words and Video can use available source imagery. In real-estate projects, listing photographs can be classified by room or scene type and sequenced into a property tour.
When the source does not contain the image a scene needs, the system can generate a new visual from a scene-specific prompt. Generated images can be replaced whenever a better source or user-supplied image is available.
Uploaded photographs, approved library images, interface captures, and local visual assets can be assigned to specific scenes and prioritized over generated imagery. Real-estate projects can also use a photos-only workflow without AI-generated image fill.
Words and Video turns the visual plan into a finished production. Landscape and portrait outputs are supported across different formats.
Language, orientation, visual treatment, and source methods depend on the selected format.
Start with a supported URL, source text, finished script, uploaded images, or a combination.
Use suitable source images, supply approved imagery, allow AI generation to fill gaps, or use a photos-only workflow where available.
The system prepares the scene sequence, narration, captions, and render. Individual visuals, wording, pacing, and audio can then be revised without rebuilding the entire project.
Words and Video uses dedicated workflows for different kinds of video rather than forcing every project through the same template.
Each format has its own visual language, input rules, narration, pacing, and output configuration.
Hebrew or English civic-information videos for public announcements, local updates, municipal communication, and place-based stories.
Hebrew or English property videos built from supported listing pages and optional uploaded photography. Listing images can be classified, matched to scene types, and sequenced into a property tour.
Hebrew or English portrait videos built from prepared scripts and scene-by-scene visual planning. They can combine curated, generated, and user-approved visuals.
Hebrew or English roundup videos combining imagery from supported articles with generated visuals when source material cannot cover a scene.
Hebrew market-news videos built from news inputs, source material, and finance-specific visual and narration rules.
Hebrew or English landscape videos for a single story, concept, product, or workflow. They can combine interface captures, supplied assets, generated visuals, and scripted narration.
English landscape formats for structured briefings with their own visual identity, narration, title cards, and pacing rules.
A Hebrew football and basketball roundup format currently available as a pilot workflow.
Finished output, with the source type and language it was produced from.
Listing photographs classified by room type and sequenced into a walkthrough.
A town described through its public spaces, from the ballpark to the food hall.
Expedition photographs where they exist, generated scenes and data views where they do not.
A finished script planned scene by scene as a portrait essay.
Four starting points in one row: a listing page, a prepared story, supplied field photography, and a finished script — two languages, one workflow.
Several stories in one landscape edit, with generated visuals covering the scenes the articles could not.
Currently available in Hebrew and English, depending on the selected format.
Words and Video was built by Yoav Gerber, a writer, translator, and independent content creator looking for a more repeatable way to produce video without maintaining a separate tool and handoff for every stage.
The system brings source collection, visual planning, image generation, narration, captions, and rendering into configurable workflows.
Each supported format is developed around a real production use case and improved through actual use.
Send a URL, brief, script, or sample images. We'll identify the workflow that fits, what can be reused from the source material, and where supplied or generated visuals would be needed.
enquiries@wordsandmusic.techThank you — your mail client should open with the details. If it does not, write to enquiries@wordsandmusic.tech.
