Concurrent Submit | zhenzhen-vidu-q3-video-lowprice

Vidu Q3 is the current-generation model, and this node is the "everything" version: it exposes thirteen model choices covering text-to-video, image-to-video, start-end, reference-to-video, and an ad lane - all through the pack's low-price domestic channel. If you're on the 贞贞的平价AI小屋 (api.seedance.nz) account, this is the node that gets you the whole Q3 family at roughly three-quarters of the overseas price. The concurrent wrapper then lets you batch those runs ten deep in the shared video pool, which is how you actually compare pro vs turbo vs pro-fast on the same prompt without waiting out five sequential generations.

Inputs that matter

• model - the thirteen choices: vidu-q3-pro-t2v, vidu-q3-turbo-t2v, vidu-q3-pro-fast-t2v, vidu-q3-pro-i2v, vidu-q3-turbo-i2v, vidu-q3-pro-fast-i2v, vidu-q3-pro-start-end, vidu-q3-turbo-start-end, vidu-q3-pro-fast-start-end, vidu-q3-r2v, vidu-q3-mix-r2v, vidu-q3-ad-r2v, and one more in the sample list. Reading the naming: pro = quality, turbo = speed, fast = the accelerated tier, t2v/i2v/start-end/r2v = the mode. This single dropdown replaces three separate nodes.
• prompt - the scene description.
• seconds - 4 through 15 seconds, default 4.
• ratio - adaptive, 16:9, 4:3, 1:1, 3:4, 9:16, 21:9.
• resolution - default, 720p, or 1080p.
• image1–image9 - references for the i2v/start-end/r2v modes; how many you use depends on the model you picked.
• seed - defaults to -1 (randomize), unlike most nodes' 0.
• api_config, skip_error - the low-price channel config, from the zhenzhen-seedance2-low-price-api-settings node or the SEEDANCE_API_KEY environment variable.

Output: one COMFLY_VIDEO_FUTURE task.

How it works

Standard ComflyConcurrent.py wrapper: validate, push the original Comfly_vidu_q3_video_lowprice call onto the bounded video executor (ten workers), return the future immediately. The Concurrent Collect Videos node waits on all tasks, keeps your slot order, and reports per-slot status; failure_mode: placeholder turns failures into blank clips.

Setup

Install the pack once - ComfyUI Manager (search "Comfyui-zhenzhen"), or:

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

Key it via the low-price settings node or SEEDANCE_API_KEY - the domestic channel is the discount that justifies this node's existence.

Gotchas

The model list is the trap: "pro-fast" and "pro" and "turbo" are genuinely different price/quality tiers, and picking blind gets you either slow runs or visibly cheaper output. If you're comparing, run the same prompt across tiers in parallel - that's what the concurrency is for. The r2v and start-end modes want their reference images connected; forget them and the job just fails or drifts. And don't confuse this low-price key with your main Zhenzhen key - the settings node is what makes it work. Timeouts aren't guaranteed failures; check the async task page before you rage-re-run.