Oyama-AI-Video-Studio-Setup-1.0.0-alpha.72

Oyama AI Video Studio 1.0.0-alpha.72
This update adds new ways to continue, edit, and assemble locally generated video, alongside fixes for ComfyUI media handling.
New workspaces

• Continue builds a sequence of H3 beats from an original clip. Each completed beat can supply the motion context for the next, with a sequence player, source selection, prompt editing, camera controls, and clip exports.
• LTX Ripple carries an edited first frame through a source video. It supports custom edit lengths, overlapping chunks for longer clips, resolution controls, and live previews.
• Photo Edit uses FireRed to edit an extracted frame and send it directly to Ripple as a replacement frame.
• MiniMax Music 3 adds long-form local music generation with lyrics, arrangement controls, and in-app playback.
  Improvements
• Refined the workspace navigation, responsive layouts, media previews, and project persistence.
• Added same-resolution H3 refinement and more reliable render progress and cancellation feedback.
• Improved large-video preparation, ComfyUI output-folder detection, and recovery of saved jobs.
• Fixed Continue source handoffs and Ripple replacement-frame, preview, and output states.
• Updated Music 3’s MP3 save graph to use the format.quality input required by current ComfyUI.
  Upgrade note: Ripple, Photo Edit, and H3 motion continuation require their corresponding ComfyUI models and custom nodes. Use Settings → Test connection and Rescan models after updating ComfyUI or installing nodes.

V0.12.86

Full Changelog: 0.12.25...0.12.86

V0.12.25

New Blue Theme - General Fixes. and more features.

New Theme

I have not gotten to test the linux version yet... but let me know what issues you have in discussions. and i will fix them.

V0.12.22

Full Changelog: V0.12.0...0.12.22

V0.12.0

What’s New
MiniMax H3 Encoding

• Added optional qwen3vl_32b_minimax_h3_int8_convrot.safetensors text encoder.
• Fast NVFP4-AWQ remains the default.
• New “Slower · better encoding” quality option is available in Create → Advanced controls and Settings → Generation defaults.
• Quality mode will not silently fall back if the INT8 encoder is missing.
  Prompt Builder
• Added automatic reference summaries with visual source chips.
• Reference assignments can be reviewed in a dedicated modal.
• Added copyable prompt output and improved prompt insertion tools.
• Ollama prompt assistance now produces concise, timeline-aware additions that avoid repeating existing instructions.
  Video Generation
• Added MiniMax H3 Turbo Ref2VA 8-Step v1.0 support.
• Added exact continuation workflow using the previous render’s final frame.
• Continuations preserve canvas size, sampling settings, motion settings, and LoRA configuration.
• Removed the redundant manual final-frame extraction flow for continuation work.
  Rendering and Progress
• Added live video render progress for LTX 2.5 location-set generation.
• Progress now includes current stage, sampler steps, and percentage completion.
• Improved queue, cancellation, and render status feedback.
  Chat Workspace
• Enlarged the AI chat workspace.
• Added pop-out chat windows.
• Added always-on-top support.
• Added dock/restore controls for returning the chat to the main window.
  Reference and Location Workflows
• Improved character, wardrobe, location, and reference-media assignment.
• Added clearer locked reference instructions with editable source management through modal controls.
• Improved responsive layouts for desktop, resized windows, and mobile views.
  Reliability
• Added persistence for encoder preferences and generation defaults.
• Expanded model detection and validation.
• Added regression coverage for official H3 model selections and Turbo workflows.
• Verified with linting, TypeScript checks, production build, and workflow tests.

Full Changelog: https://github.com/jamesk9526/MINIMAX-DESKTOP/commits/V0.12.0