[Frontend] Add MiniMax-H3 reference-to-video workflow

[LinzeShi]

Summary

Add the MiniMax-H3 Ref2VA reference-to-video workflow for ComfyUI.

The workflow supports up to 9 image references, 3 video references, and 3 audio references through the expanded References node, with automatic ref2va routing to the remote vLLM-Omni video API.

Validation

• python -m pytest tests/e2e/features/comfyui/test_h3_reference_workflow.py -q
• Validated end-to-end with a real MiniMax-H3 Ref2VA service on 2x NVIDIA A100 80GB GPUs.
• Generated a 1344x768, 24 FPS, 124-frame video from one image reference.
• Verified the saved MP4 contains AAC stereo audio at 32 kHz.

Dependencies

• Requires the expanded reference input support from [Frontend] Add MiniMax-H3 references and Ref2VA workflow #7483.
• Audio-preserving output requires [Bugfix][ComfyUI] Fix audio track silently dropped in bytes_to_video #6782 or [Bugfix][Frontend] Preserve generated audio in ComfyUI video output #5905.

The generated workflow screenshot and real-model result will be attached to this PR.

[chatgpt-codex-connector]

Codex Review Summary

This comment shows the latest Codex review activity on this pull request.

Review	Status	Commit	Review trigger
📝 Code Review	✅ Completed 2026-09-15T03:10:14.429321Z	b591845	PR opened

ℹ️ About Codex in GitHub

Your team has set up Codex to review pull requests in this repo. Reviews are triggered when you

• Open a pull request for review
• Mark a draft as ready
• Comment "@codex review" or "@codex security review".

Codex reacts with 👀 while any review is running, comments if it has suggestions, and reacts with 👍 once all reviews finish with no findings.

[chatgpt-codex-connector]

💡 Codex Review

Here are some automated review suggestions for this pull request.

Reviewed commit: b5918457df

ℹ️ About Codex in GitHub

Your team has set up Codex to review pull requests in this repo. Reviews are triggered when you

• Open a pull request for review
• Mark a draft as ready
• Comment "@codex review".

If Codex has suggestions, it will comment; otherwise it will react with 👍.

Codex can also answer questions or update the PR. Try commenting "@codex address that feedback".

[chatgpt-codex-connector]

Document soundtrack-aware audio tag numbering

When a connected reference video contains audio and a standalone audio_* slot is also connected, this instruction maps the standalone clip to the wrong prompt tag. MiniMax-H3 assigns <Audio 1>, etc. to embedded video soundtracks first, then numbers standalone audio references afterward (pipeline_minimax_h3.py:2463-2470), so a user following this guidance may condition on the wrong audio source. Explain that soundtrack-bearing videos consume audio indices before connected audio slots.

Useful? React with 👍 / 👎.

[LinzeShi]

Closing this duplicate draft PR; WF-03 is being folded into #7483 so the Node-02 and WF-03 commits can be reviewed together.