AI Community

Text-to-image and video generation have a limitation: words alone make it hard to specify exactly where the camera should be and how objects should relate to one another. “A little lower,” “closer to the foreground,” or “the car at an angle to the building” can be interpreted differently every time.

With 3D Animatic, you can position the camera and objects yourself, preview the composition in the viewport, and prepare a visual reference with the framing and movement you want.

Choose a viewpoint, adjust the field of view, change the distances between objects, and check exactly what fits in the frame. If the shot needs motion, set it up on the timeline and play it back.

Build a scene from a description or an image

Start with an empty scene, a text prompt, or an image. The built-in AI agent helps block out a scene using simple shapes: a room, a street with cars, a courtyard, a natural environment, or a product composition.

Attach a reference and describe the task. The agent will approximate the main shapes and object placement. You can watch the scene take shape gradually in the viewport.

You do not need a detailed model of every object. Recognizable silhouettes, suitable proportions, and the right distances are enough to help plan the shot.

Position objects exactly where you need them

Once the scene is built, you can manually move, rotate, scale, or recolor any object. Bring a car closer to the camera, turn a table, increase the gap between buildings, or remove an object that blocks an important part of the frame.

You can also ask the agent to make changes:

• “Move the car to the right and keep the center of the frame clear.”
• “Turn the sofa toward the window.”
• “Place the trees farther from the camera.”
• “Remove the front wall.”
• “Make the passage between the objects wider.”

After you send a reference, it remains in the chat history and stays in context for subsequent edits. Upload an image from your device or drag one into the chat from Neylora's mini-gallery.

You can stop the agent while it is building, and undo its changes with a single Undo.

Set the camera for your shot

The camera is central to the animatic workflow. Choose its position and viewing direction, adjust the field of view, and evaluate the composition through the camera view.

Check how a low viewpoint changes the shot, how large an object appears, what sits in the foreground, and how objects overlap.

Instead of repeatedly trying to explain an angle in words, you have a scene in which you can set the composition directly.

Animate the camera and objects

A keyframe timeline lets you set objects and the camera in motion. Define their starting positions, intermediate points, and final positions over time.

Plan a camera move along a street, a push-in toward a subject, a moving car, or an object emerging from behind an obstacle.

Play the animation to check timing and how the elements move in relation to one another. Visible motion paths help you inspect an object's route through the scene and adjust it.

This is especially useful when you need to establish where the camera starts, where it ends, and where the objects should be relative to it.

Refine the geometry when needed

Beyond positioning ready-made shapes, you have modeling tools for working with faces and edges, including loop cuts, extrusion, inset, and bevel.

Use them to adjust a shape's silhouette or build a specific structure, such as a wall with an opening. Block out the scene quickly with the agent, then manually refine the geometry that affects the shot.

Organize the scene and save the result

Group timeline objects into folders, reorder them, and hide or lock them. Keeping the environment separate from moving objects helps keep the workspace manageable.

Collapsed folders display keyframe markers for their contents, so you can see where animation occurs without expanding every track.

Projects are saved in Neylora folders. Within a project, you can create multiple scenes to explore different camera setups and object arrangements. Export the finished animation as a video directly to your reference library.

When reconstructing a space from a single image, dimensions and hidden areas are only approximate. Review the result and adjust it where necessary. Simple geometry helps you focus on what matters: camera placement, composition, and object movement.

You can now refine an image right where you generated it. Open an image from your gallery or upload one from your computer.

Create collages, add text and shapes, draw, rotate and flip objects, and adjust opacity and layer order—all on a single canvas.

For AI editing, describe what you want to change. Without a mask, your instructions apply to the whole image; use the brush to mark a specific area. Arrows and shapes can help show where and how to make the edit. You can also add reference images and compare the result with the original.

In the free Neylora Studio app, the editor also supports filters, color correction and effects—for a selected image or all images on the canvas.

Nano Banana 2.1 is Google's updated image generation and editing model. The release may have been quiet, but the changes are worth a look: closer instruction following, better continuity across successive edits, and improved control over local changes.

Nano Banana 2.1 is already available in Neylora.

Your generated video looks just right—except for the background music getting in the way of your edit? You can now remove it directly in Studio while keeping speech and sound effects.

The new Remove music tool works with video clips and standalone audio clips, including audio you have already detached from a video.

Two processing modes

• Remove music isolates the music and subtracts it from the original audio. You get a single No music track containing speech and effects.
• Split into three tracks creates separate tracks for voice, sound effects and music. This mode usually provides cleaner separation and more flexibility when editing. The music track is added muted.

Results depend on the clip: the sounds it contains, how loud the music is and how closely it overlaps with speech. Try both modes to see which works best for your footage.

Ready to edit, right on your timeline

The processed tracks are added separately, and the original audio is muted. The clip's trim, position and speed are preserved, so you do not need to sync the audio to the video again.

The resulting WAV files also appear under Files on PC. Their names include the source filename, making it easier to identify the voice or music-free version. You can reuse them in the project or take them into another editor.

Studio processes the source's full audio, so you can keep adjusting the clip boundaries after separation. Adding an already saved result again does not incur another processing charge.

Try it out

Select a clip and click the music-note icon in the toolbar, or choose Remove music from its right-click menu. Pick a mode, check the credit cost and start processing. You can follow the progress, or hide the window and carry on editing.

There is also a headphones button in the toolbar to quickly mute or unmute selected clips. The keyboard shortcut is M.

Draft mode for Seedance 2.5 is now available in Neylora, helping you test ideas with fewer credits.

Start with a 480p draft to check the scene and motion before generating the final version. Happy with the result? Click “To 1080p” on the draft card. The model recreates the video at higher quality from that same draft, keeping the shot sequence and movement—not simply upscaling the low-resolution file.

Draft mode also supports photorealistic faces.

The draft and final version are charged separately. The cost of the final version is shown on the button before you start.

This lets you explore and compare ideas without spending credits on a high-resolution render every time.

Turn text into speech, create multi-speaker dialogues, and control the emotion in every line.

Choose voices from the library, clone your own voice from an audio recording, or create a new voice from a description. Save your favorite voices in projects, download the finished voiceover, or add it to your gallery and references.

Wan 3 by Alibaba is now available in two versions: Pro and Fast (Prime). Pro is the standard Wan 3 model; Fast is an accelerated version for quick iterations.

Теперь готовить промпты и сценарии можно прямо в Нейлоре. Опишите идею, прикрепите референс или принесите черновик — писатель поможет собрать текст и доработать его в диалоге.

Чат и документ находятся рядом: обсуждение остаётся в переписке, а готовый текст — в редакторе. Можно попросить изменить отдельный фрагмент, разбить сценарий на сцены и хранить материалы по проектам.

Для работы с конкретными нейросетями есть навыки — инструкции, которые помогают учитывать особенности модели при написании промпта. Можно использовать системные навыки или создать свои под привычные задачи и стиль. Язык готового текста выбирается отдельно.

Писатель нужен, чтобы быстрее пройти путь от «есть задумка» до промпта или сценария, с которым уже можно работать. А готовый промпт можно сразу отправить на генерацию.

Мы выпустили собственное приложение, в котором AI-генерация и полноценный видеомонтаж объединены в одном месте.

В Neylora Studio можно:

— генерировать видео и изображения с помощью AI;

— добавлять материалы на таймлайн и собирать ролики;

— обрезать и компоновать фрагменты;

— работать со звуком;

— настраивать цветокоррекцию;

— применять эффекты и маски;

— создавать анимации с помощью ключевых кадров;

— сохранять проекты и экспортировать готовые видео.

Больше не нужно постоянно переключаться между генераторами и монтажными программами: создавайте контент и сразу превращайте его в готовый ролик внутри одного приложения.

Neylora Studio уже доступна бесплатно для Windows и macOS.