minimax_wf
An open-source workflow for MiniMax AI video and image generation, hosted on GitHub. It supports node-based operation in ComfyUI and integrates capabilities including SAM segmentation models and audio control.
Timeline (last 90 days)
A developer released the V5 version update of the MiniMax workflow (minimax_wf).
MiniMax workflow V5 introduces SAM 3.1 latent space masking technology for guided character replacement.
V5 adds a block audio feature that forces MiniMax to change the generated voice every time.
The new resolution selector in V5 supports two decimal places of precision, allowing resolution values such as .98 MP to be selected.
V5 adds group bypass controls, allowing users to bypass certain node groups to avoid errors.
Developer roycho87 published an open-source project on GitHub featuring an optimized MiniMax ComfyUI workflow.
The workflow has an automatic citation routing mechanism that dynamically adjusts internal node connections based on the type and number of citations enabled by the user.
The workflow introduced a centralized control panel that consolidates all key settings into a single control center.
The workflow supports up to 6 image references, 3 audio references, and 2 video references, which can be combined in a single generation.
The workflow supports running on devices with limited VRAM through sparse attention and a low-VRAM mode (chunked processing and attention mechanism optimization).
1 more timeline events
All Facts (10)
The workflow has an automatic citation routing mechanism that dynamically adjusts internal node connections based on the type and number of citations enabled by the user.
50%UnverifiedThe workflow introduced a centralized control panel that consolidates all key settings into a single control center.
50%UnverifiedThe workflow supports up to 6 image references, 3 audio references, and 2 video references, which can be combined in a single generation.
50%UnverifiedThe workflow supports running on devices with limited VRAM through sparse attention and a low-VRAM mode (chunked processing and attention mechanism optimization).
50%UnverifiedThe workflow supports multi-LoRA model loading and FPS control functionality.
50%UnverifiedV5 adds group bypass controls, allowing users to bypass certain node groups to avoid errors.
50%UnverifiedMiniMax workflow V5 introduces SAM 3.1 latent space masking technology for guided character replacement.
50%UnverifiedV5 adds a block audio feature that forces MiniMax to change the generated voice every time.
50%UnverifiedThe new resolution selector in V5 supports two decimal places of precision, allowing resolution values such as .98 MP to be selected.
50%UnverifiedDeveloper roycho87 published an open-source project on GitHub featuring an optimized MiniMax ComfyUI workflow.
50%