A creative team
with a memory.
Nolyr connects the why of a campaign to the images and video that ship. One Agent leads, a team of AI models argues the idea, and every generation waits for your approval.
It thinks from several sources.
The lead Agent talks to you. For big decisions it asks the rest of the team for independent takes — a strategist, a copywriter, an art director and a media buyer, each running on a different AI provider — then keeps the best ideas and says whose they were.
- Replies in your language, Arabic dialects included
- Keeps the campaign brief current as it learns
- Chooses the model per job and explains why
Frames first. Then motion.
For a video ad the Agent writes a storyboard: a hook, a reveal and a payoff, each scene with its own shot. Frames are made as images first — cheap to redo — and only approved frames become video.
- One take: the model directs one continuous video from all frames
- Per scene: a clip for each frame, cut together with crossfades
- Use your own product photo for any scene
- 3D looks — render, glass, clay — and orbit, push-in, fly-through shots
Words are typeset, not generated.
Models are great at motion and bad at spelling. So Nolyr keeps text out of the generation and lays captions, the price card, your logo and footer on top at the end — exact, sharp, and right-to-left when it should be.
- Logo, colours, look references and footer saved per campaign
- End card with headline, price and call to action
- 9:16, 4:5 and 1:1 exports, ready to post
Every model, when you want the wheel.
Skip the Agent whenever you like. The Studio lists the strongest image and video models — Kling, Seedance, Wan, MiniMax, Grok, Ideogram, Recraft and more — with each model's own settings, a live cost estimate and one library for everything you've made.
The same face, every time.
Train a character from photos of one person and the Agent can cast them across a whole campaign — creator-style ads, product shots, feed posts — without the face drifting between images.
Train once from 1–100 photos of one person, then cast them anywhere.
Nothing runs until you say yes.
The Agent proposes; you decide. Every image, storyboard step and video shows the model, the settings and the cost before anything is spent. Big runs ask for one more confirmation with the total.
See pricing