PixVerse V5.6

PixVerse V5.6 is the newest PixVerse text-to-video node in the Runware pack, and it's the one that finally answers the two questions everyone has about AI video: can it reason about the scene, and can it bring its own sound. The node adds a providerSettings.pixverse.thinking toggle - enabled, disabled, or auto - that lets the model spend a reasoning pass on the prompt before it renders. And it adds providerSettings.pixverse.audio, a straight on/off for generating an audio track with the video. Native audio in AI video is still the capability gap open models haven't closed (the KB's closed-source-models essay calls it out explicitly), so this is the feature that makes V5.6 worth a look over its siblings.

The thinking mode is the sleeper, though. Prompts that trip up older video models - "two people crossing, one left to right, one right to left" - are exactly the multi-part instructions a reasoning pass handles before the pixels get drawn.

What you actually set

• positivePrompt - required. With thinking enabled, be detailed; the reasoning pass has something to chew on.
• frameImages - IMAGE socket for image-to-video.
• providerSettings.pixverse.audio - on/off for a generated soundtrack. This is the headline new capability.
• providerSettings.pixverse.thinking - enabled, disabled, or auto (default). Auto lets Runware decide; enabled is for complex prompts, disabled for speed.
• duration - note the shape changed: it's now a gated toggle with a duration_value float underneath, so you get finer control than the old 5-or-8 choice.
• size - the usual resolution presets, default 720p (16:9).
• providerSettings.pixverse.style - anime, 3d_animation, clay, comic, cyberpunk.
• negativePrompt, seed, numberResults - the usuals.

Output: one video socket, type VIDEO.

Install

The one-time pack setup: ComfyUI Manager → search Runware → install → restart, or

cd ComfyUI/custom_nodes
git clone https://github.com/Runware/ComfyUI-Runware
pip install -r ComfyUI-Runware/requirements.txt

API key into ComfyUI Settings → Runware API key or RUNWARE_API_KEY. Light deps, no downloads.

Where people get burned

• Thinking costs. thinking: enabled adds a reasoning pass before generation, which is time and money on top of an already premium model. Keep it on auto unless a prompt genuinely needs it.
• Audio isn't guaranteed quality. The generated track is real, but it's a model guess at sound design. For dialogue or precise sync you're still better off adding audio in post.
• Newest ≠ cheapest. V5.6 sits at the top of the PixVerse price ladder alongside V5. The title bar shows the per-run cost for a reason.
• safety.mode defaults to none. Same opt-in story as the rest of the video nodes.

The verdict: V5.6 is the most complete PixVerse node - thinking for tricky prompts, audio for sound-on shorts, finer duration control. If you want one PixVerse node that does the most, this is it.