Concurrent Submit | zhenzhen-qwen-image-3.0-lowprice

The newest Qwen image node in the pack, and the one with the most model choices. ComflyConcurrent_Comfly_qwen_image_3_0_lowprice_Submit wraps Alibaba's Qwen Image 3.0 family - and "family" is doing real work here, because the model dropdown has eight entries: qwen-image-3.0-t2i and -i2i, the same pair with -pro- in front, and four -global- variants (standard and pro, t2i and i2i). Text-to-image, image editing, budget tier, pro tier, domestic routing, global routing - one node, eight lanes.

The "lowprice" in the name means this one points at the domestic shop (api.seedance.nz), the non-profit-ish site that's ~7.5–8折 cheaper than the overseas store. That's why the node takes an api_config input of type ZHENZHEN_SEEDANCE2_CONFIG - you build it with the Comfly Seedance2 Low Price Settings node (Comfly_seedance2_low_price_settings), which just carries a base_url and api_key. If you already have that settings node in a Seedance workflow, this plugs straight in.

Concurrent-wise it's business as usual: shared image pool (30 workers), output is a COMFLY_IMAGE_FUTURE task, and you collect with Concurrent Collect Images (ComflyConcurrent_Image_Await) which waits on up to 30 tasks and returns image_1…image_30 in order.

The inputs that matter

• model - the eight-way dropdown above. Pick global variants if you're outside China; the domestic lanes are optimized for mainland routing.
• prompt / negative_prompt - both multiline, and Qwen responds well to a real negative prompt.
• prompt_extend - on by default; lets the model expand your prompt. Turn it off when you want exactly what you typed.
• sizing_mode - auto (follows ratio), ratio (pick a ratio), or custom_size (type your own).
• resolution - 1k or 2k; ratio - 1:1, 16:9, 9:16, 21:9 and the rest; custom_size - e.g. 1024*1024.
• n - 1 to 6 images per call. Six at 2K is a real bill; spend consciously.
• image1 … image3 - up to three references for the i2i models.
• seed - -1 (random) to 2147483647.

Install

The manual route, same as every node in this pack:

cd ComfyUI/custom_nodes
git clone https://github.com/T8mars/Comfyui-zhenzhen

Or ComfyUI Manager → search Comfyui-zhenzhen → restart. Then build the api_config with the Seedance2 Low Price Settings node and your domestic-site token.

Gotchas

• The domestic site is region-optimized. If you're not on a China-friendly connection, the global model variants and the overseas token are the safer combo. The README's network advice - 443 errors mean your route to their server is the problem, not your node - applies.
• Prompt extension changes results subtly. If a batch comes back "not what I asked for," check prompt_extend before you blame the model.
• n × resolution × pro = sticker shock. The low price is per-image; six pro images at 2K still add up. Watch the async-task log if the number surprises you.