Vitent
Changelog
Verified product releases, improvements, and fixes for the English Vitent experience. Major releases include a full detail page; smaller changes stay grouped in this timeline.
Vitent 1.0: Agent and Drama creation modes
New capabilities
- Agent mode: Describe the video you want, answer the Agent's questions, review the proposed assets and video plan, and generate a finished result through a guided workflow.
- Drama mode: Organize scripts, episodes, scenes, shots, reusable assets, storyboards, and generation work in a structured production workspace.
- Creation templates: Start from purpose-built workflows for common tasks such as commercials, product videos, explainers, reference remakes, and creator content.
- Inspiration library: Browse a broad collection of completed examples by use case, then adapt a relevant format, pacing pattern, or creative direction with your own approved content.
Landing page and canvas video fixes
Experience improvements
- Landing page presentation: Adjusted video playback duration and button copy.
- One-click video import: Corrected aspect-ratio handling when importing video to the canvas.
- Current-frame extraction: Improved the logic used to extract the current frame from a video.
Checkout and homepage showcase improvements
Experience improvements
- Checkout dialog redesign: Reworked the credit-purchase dialog so checkout choices and the payment path are easier to understand. Availability varies by region.
- Homepage showcase refresh: Updated the featured cases so visitors can understand what creators made with Vitent and how each workflow was used.
Open registration and account access improvements
New capabilities
- Open registration: Anyone can now create a Vitent account without an invitation code.
- Login and registration redesign: Simplified account creation, sign-in, and the route into the product.
Smart cutout, read-only canvases, and text model selection
New capabilities
- Smart cutout for image nodes: Isolate the main subject of an image for further editing, compositing, or generation.
- Read-only canvas mode: Share a canvas for preview and review while reducing the chance of accidental edits.
- Model selection for text nodes: Choose an available GPT-series model for the text task at hand.
Experience improvements
- Clearer uploaded-content titles: Improved how canvas nodes label uploaded material.
- Canvas performance: Improved node rendering and interactions on complex canvases.
Richer Agent history and faster storyboard navigation
New capabilities
- Media nodes in Agent conversations: Review video and image nodes directly alongside the related discussion.
- Longer Agent history: Load earlier messages by scrolling upward through a conversation.
- Linked prompt duration: Select duration while referencing content and keep it aligned with the generation parameters.
Experience improvements
- Missing-image navigation: Jump from an incomplete storyboard to the relevant asset details and add the missing material.
Image splitting, voice samples, and smoother video previews
New capabilities
- Split images into grids: Turn one canvas image into smaller asset units for editing and generation.
- Custom voice sample rules: Clarified the minimum duration required for a custom voice sample.
Experience improvements
- Timeline full-screen playback: Fixed an issue that closed full screen when video preview resumed automatically.
- Canvas video feedback: Improved video loading and generation-progress interactions on the canvas.
Credit estimates and generation parameter validation
New capabilities
- Dynamic credit estimates: See an estimated credit cost before submitting a generation task.
- Generation parameter validation: Catch incompatible model settings before generation to reduce avoidable failures and rework.
Creator Program and credit-earning entry points
New capabilities
- Creator Program entry point: Added a direct route to current creator activities and incentives.
- Earn Credits entry point: Added a homepage route to the available ways to earn credits.
Image editing, Virtual Studio preview, and review guidance
New capabilities
- Crop and rotate images: Adjust generated and uploaded material before reuse.
- Improved annotations: Add clearer visual notes for review, revision, and collaboration.
- One-click Virtual Studio preview: Review scene layout and camera framing more quickly.
Experience improvements
- Default visual styles: Set a default style to keep new projects visually consistent.
- Human-subject review guidance: Follow clearer guidance when a generated person requires further review.
Canvas references and task-to-canvas navigation
New capabilities
- @ references on Free Canvas: Reference nodes from prompts to connect source material, text, and generation tasks.
- Return to a task's canvas: Open the relevant canvas directly from a generated video in task notifications.
Storyboards, task feedback, and video delivery improvements
New capabilities
- Video upscaling: Increase the resolution of a finished video for clearer review and delivery.
- Referral rewards: Invite collaborators through the available referral program.
Experience improvements
- Storyboard redesign: Improved storyboard viewing, editing, and organization.
- Task notification redesign: Consolidated progress and status feedback for asynchronous generation tasks.
- Agent task progress: Display progress for supported non-generation tasks handled by the Agent.
Canvas and storyboard grouping
New capabilities
- Canvas groups: Organize related nodes into standard groups or storyboard groups on complex canvases.
- Storyboard groups: Keep the nodes for a clip or shot together for easier review, movement, and collaboration.
Virtual Studio Agent, character interactions, and end-frame review
New capabilities
- Agent support in Virtual Studio: Use the Agent to assist with scene edits, camera changes, and production tasks.
- Expanded studio editing: Build spaces, arrange elements, and manage camera setups with more control.
- Character interactions: Stage character relationships and interactions within a controlled scene.
Experience improvements
- End-frame review: Send the final frame of a video into human review to support continuity and quality control.
More image models and clearer pricing
New capabilities
- Additional image models: Choose from supported Midjourney-series models during image generation.
- Updated image-generation pricing: Revised credit costs and clarified the applicable billing rules.
Reliability and polish
- Reliability fixes: Resolved known issues across the generation workflow.
Global references and multimodal Agent understanding
New capabilities
- Global @ references: Open the reference picker with @ and use Shift to select multiple project items.
- Quick references from documents: Select the @ control in the document sidebar to cite content in a prompt.
- Reference menu shortcut: Open the global @ menu with a keyboard shortcut.
- Reference selected document text: Cite a selected passage directly when working with scripts, settings, or notes.
Experience improvements
- Workbench guidance: Let the Agent suggest a jump to the relevant production workspace when the next step requires it.
- Multimodal understanding: Give the Agent more context from text and media used in the project.
Shared cases, DOC uploads, and a unified video player
New capabilities
- Case sharing: Publish a viewable case so collaborators and prospective users can understand a finished workflow.
- DOC uploads: Bring scripts, settings, and other creative material from DOC files into a project.
- Unified video player: Use consistent playback controls across preview and review surfaces.
Experience improvements
- Richer generation confirmation: Review prompts, parameters, and risk notices together before submitting a video task.
- Simplified prompt component: Present the title and description with less visual noise.
- More reliable HITL notices: Improved how the interface renders operation notices from human-in-the-loop cards.
Reliability and polish
- Clip revision refresh: Fixed state refresh after a partial Clip revision.
Reusable voices, audio nodes, and project previews
New capabilities
- Voices linked to assets: Associate a character or narration asset with a reusable voice.
- Voice library: Manage and select character and narration voices in one place.
- Custom voices: Create a reusable project voice for more consistent audio across a series.
Experience improvements
- Audio generation on canvas: Generate audio nodes alongside images, text, and video tasks.
- Shared project preview: Let reviewers view a project through a shareable preview.
Reliability and polish
- Full-episode preview fix: Improved the reliability of episode preview links.
Rich-text storyboard prompts and Agent editing
New capabilities
- Rich-text storyboard prompts: Read and revise structured storyboard prompts more easily.
- Storyboard editing with Agent: Add, remove, or revise part of a storyboard through the Agent.
Experience improvements
- Adaptive cover images: Improved how project and episode covers fit across different image sizes.
- Clearer team setup: Moved team creation into project management so the control is easier to find.
- Simpler credit-purchase page: Reduced secondary billing copy so the purchase flow is easier to scan.
Reliability and polish
- Stable Agent streaming: Fixed interface movement while Agent responses stream.
- Virtual Studio camera controls: Fixed an interaction issue between the camera and viewfinder.
Candidate assets, public sharing, and review downloads
New capabilities
- Asset forms and candidates: Generate or upload multiple candidates, choose the active image, and copy, move, or remove alternatives.
- Expanded public asset sharing: Show and download multiple asset forms from a shared view.
- Review package download: Download the selected video set as one package for delivery, review, or archiving.
Experience improvements
- Candidate history: Save generated results to the relevant asset form and support multi-form targets.
- Source information on import: Preserve media and origin details when importing images into Timeline or Asset Canvas.
- Expanded Agent asset controls: Let the Agent work with candidate images and more detailed Virtual Studio operations.
Reliability and polish
- Virtual Studio performance: Improved model labels, screenshots, camera behavior, and scene navigation.
Episode covers and Virtual Studio workflow improvements
New capabilities
- Custom episode covers: Give each episode a distinct cover for faster recognition.
- Flexible credit purchase: Enter a custom credit amount where the option is available.
- Virtual Studio shortcuts: Use undo, redo, and delete from the keyboard.
Experience improvements
- Preserved image proportions: Keep the original aspect ratio when adding an image through a canvas link.
- Clearer studio feedback: Improved panorama and screenshot-history feedback during scene review.
Reliability and polish
- Video adoption fix: Corrected the copy behavior when adopting a canvas video result.
- Studio tag synchronization: Improved consistency of Virtual Studio tags.
- Camera and perspective fixes: Resolved screenshot and view-switching issues when models overlap.
3D Virtual Studio preview
New capabilities
- 3D Virtual Studio: Build a spatial scene layout and compose camera views before video generation.
- Improved candidate assets: Keep more image alternatives for an asset and choose the best fit for the next production step.
Custom visual styles and smoother asset generation
New capabilities
- Custom visual styles: Create, edit, and remove styles, or derive a style description from reference images.
- Style management: Expanded title and description limits and improved list feedback.
- Clearer batch asset details: Show model, dimensions, aspect ratio, and other context for Agent-generated asset batches.
Experience improvements
- Homepage project creation: Improved input guidance and allowed an empty state to start a new project directly.
- Smoother Agent output: Improved streaming Markdown and response animation.
- Batch generation flow: Keep users in context after starting a batch image task.
- Contact Sales: Added a direct route from Help Center for solution and purchasing questions.
Reliability and polish
- Persistent Agent panel size: Fixed route changes that could rerender the panel and interrupt its event stream.
- Project and workspace refresh: Improved script-list refresh and workspace interface details after script edits.
- Recent-activity sorting: Sort projects by latest activity so current work stays easier to find.
Vitent public beta
New capabilities
- Project-level organization: Coordinate plots, episodes, and reusable production material in one project context.
- Expanded Agent context: Help the Agent understand project intent, plots, assets, and earlier production decisions.
- Custom Agent specifications: Configure supported Agent behavior for a team's working style.
- Account and purchase flow: Added verification-code sign-in and supported purchasing where available.
- Seedance 2.0 Mini: Added a faster option to the video-generation workflow.
Experience improvements
- Interface redesign: Reduced visual clutter and kept key production controls easier to find.
Canvas editing and draft workflow improvements
New capabilities
- Edit assets on canvas: Revise an asset without leaving the current canvas.
- Episode asset sharing: Reuse an asset across work for the same episode.
- Draft canvas: Save early ideas before promoting them into the production workflow.
Fixes
- Duration control: Improved prompt input and added a slider for generation duration.
- Prompt write-back fix: Corrected prompt state after a canvas node finished generating.
- Unified media upload: Aligned audio and video upload behavior within projects.
- Clearer failure states: Standardized canvas error feedback and simplified the supporting interface layer.
Audio nodes, references, and prompt history
- Audio nodes: Add audio directly to the canvas for richer scene production.
- Canvas references: Link project material from prompts to improve reuse.
- Prompt history: Review the prompt and references used for a completed node.
- Single-node replacement: Replace the existing result directly when a node has only one output.
- Media understanding: Added image and video understanding to supported canvas workflows.
Seedance 2.0 human-subject video generation
- Human-subject generation: Added Seedance 2.0 support for live-action-style video workflows.
- Clearer node connections: Made compatible connection points easier to identify.
- Canvas selection: Improved selecting and manipulating nodes on the canvas.
Storyboard and Agent workflow improvements
- Storyboard empty state: Create a scene or Clip directly from an empty storyboard.
- Granular structure editing: Add, remove, or revise individual scenes, Clips, and shots.
- Markdown responses: Render Agent messages with full Markdown support.
- Response timing: Show how long the current Agent turn took.
- Copy Agent messages: Copy an Agent response for notes or reuse.
- Clearer streaming status: Changed the active state from Executing to Thinking.
- Duplicate canvas nodes: Copy and paste a node to reuse its setup.
- Stable node connections: Reworked canvas links for better reliability.
- Simplified credit access: Consolidated credit details into the balance view.
Vitent internal beta
The complete product core modules launched this time:
- Project creation and creative setup.
- Project settings.
- Project and episode management.
- Project Director Agent.
- Video production workspace.
- Script view.
- Asset view.
- Storyboard view.
- VCanvas production canvas.
- Timeline and full-episode preview.
- Account and team management.
If you have any questions or suggestions about product use, please feel free to provide feedback.