Alibaba Wan 2.7 Text To Image Pro (Official Edition)

This is the same Wan 2.7 text-to-image node as its non-Pro sibling, with one meaningful upgrade: 4K output. The base node caps at 2K; this one opens the 4096² tier that the size tooltip flags as "(Pro)". If you're making print work, hero shots, or anything that'll get looked at close up, that's the whole argument for reaching past the base node.

Same pack, same install, same cloud execution model - it's part of BizyAirPlus, SiliconFlow's "edit locally, render on their servers" layer for ComfyUI. Flip the BizyAirPlus button ON in the action bar, queue, and the node does the heavy lifting remotely. Your local machine never touches a model weight.

How it works

Identical mechanics to the base T2I node: prompt (Chinese or English, ≤5000 chars) plus size, with outputs image, request_id, and urls. The size tiers are 1K / 2K / 4K / Custom. Everything you know from the base node carries over - thinking_mode on by default for better quality at the cost of time, seed (-1 = random), color_palette JSON (3–10 hexes, ratios summing to 100%), custom_width/custom_height for the Custom size, watermark, skip_error, and api_config.

Honestly, if the schema looks copy-pasted from the base node, that's because it basically is. The difference between the two editions is the model tier behind them and the resolution ceiling it unlocks - not a pile of new knobs.

The inputs that matter

• size - the reason this node exists. 4K = 4096², and it's unavailable on the base edition.
• prompt - same as base: up to 5000 characters, bilingual.
• thinking_mode - on by default; reasoning-boosted quality, more time. Kill it for fast iterations.
• seed - same-seed stability, -1 for server-random.

Note the 4K constraint baked into the tooltip: "4K is not supported in group mode." If you're running the packed/group-mode flavor of a workflow, the Pro tier quietly falls back.

Installation

The one pack install, done once:

cd ComfyUI/custom_nodes
git clone https://github.com/siliconflow/BizyAirPlus.git
cd BizyAirPlus
python -m pip install -r requirements.txt

Or ComfyUI Manager, search BizyAirPlus. Restart, then add a bizyair.ai API key (ON-button prompt, Settings > BizyAirPlus > API Key, or BIZYAIR_API_KEY).

Common issues

• 4K isn't offered - you're on the base node. That's the whole Pro-vs-base difference.
• 4K silently downgraded - group mode forbids it; check your workflow isn't grouping this node.
• Cost creeps up - bigger sizes cost more, and thinking_mode adds time. Use 1K to iterate, save 4K for finals.
• Cloud failures - key and network first, then the ComfyUI console for the detailed error.