zhenzhen-happyhorse-1.1-lowprice

HappyHorse is ByteDance's video model that doesn't get a tenth of the Reddit coverage Kling or Seedance get - which is a shame, because it's a solid mid-range generator and, through this node, genuinely cheap. This is the happyhorse-1.1 lowprice wrapper in the Comfyui-zhenzhen pack, and it covers all three happyhorse-1.1 modes: text-to-video, image-to-video, and reference-to-video (r2v, where you hand it a batch of images describing a subject and it keeps that subject across the clip).

The pitch is the same as every "lowprice" node in this pack: you're calling ByteDance's closed API through Zhenzhen's domestic site (api.seedance.nz) at roughly 75–80% of the overseas tier's price. For a model you'd otherwise have to run yourself - and HappyHorse's weights aren't exactly consumer-card friendly - renting it per clip from a discount reseller is a reasonable trade.

How it works

Standard lowprice machinery. The api_config input is a ZHENZHEN_SEEDANCE2_CONFIG that comes from the pack's settings node, where you put the base URL and API key once. The node submits your job, polls until it's done, downloads the result, and returns a real VIDEO output plus video_url, task_id and response strings.

The model enum is small and clear: happyhorse-1.1-t2v, happyhorse-1.1-i2v, happyhorse-1.1-r2v. For i2v you connect one starting image; for r2v you connect up to nine (image1–image9) as the character/reference set.

Inputs that matter

• model - t2v / i2v / r2v. Don't wire reference images onto a t2v model; it'll ignore them.
• prompt - your scene description. HappyHorse handles both English and Chinese prompts fine.
• seconds - a string enum from 3 up (the brief's sample shows 3–14+). Default 4; shorter clips are the cheap way to iterate on motion.
• resolution - 720p or 1080p. ratio - adaptive plus the standard set.
• The image* inputs - 1 for i2v, up to 9 for r2v.

Outputs: video, video_url, task_id, response.

Installing it

Install the pack once - Manager search Comfyui-zhenzhen, or:

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

Restart, add the Comfly_seedance2_low_price_settings node, paste your key into it, and wire its api_config into this node. No model files to download - the generation is a billed API call.

Common issues

• No api_config connected - the settings node is mandatory for the lowprice family; this node has no standalone key field.
• r2v results drifting - reference-to-video keeps a subject but not forever; the pack's own guidance elsewhere is that over-stuffing references hurts consistency, so curate the strongest 2–4 images rather than all nine.
• Domestic-site reachability - api.seedance.nz is the China-facing endpoint; if it's slow or dead from your region, the overseas site is the fallback (at a higher per-call price).
• The usual suspects - empty key, 443/VPN-in-TUN-mode requirements, vendor 500s that just need a re-run, and a seed field that only controls ComfyUI's cache, not the actual output.

HappyHorse won't beat Kling or Veo on pure wow, but for "I need a clean short clip with a consistent subject and I don't want to babysit a VRAM-hungry local model," this node is quietly the right tool at the right price.