EvoLink Seedance 2.0 Mini Text-to-Video (Official)

Every EvoLink video node costs money per render, and the community's #1 complaint about Seedance via API has always been the price. So the honest headline for EvoLink Seedance 2.0 Mini Text-to-Video (Official) is that it's the cheapest video in the pack - ByteDance's Mini tier, capped at 720p, built for exactly the jobs where you refuse to pay flagship prices: batch testing, workflow debugging, prompt exploration. It's the "is this workflow even wired right" node that still gives you a real Seedance clip at the end.

What it is

The Mini surface is deliberately small:

• prompt (required) - the shot description, Chinese welcome, ~500 chars suggested.
• duration - 4–15 seconds, default 5, billed by duration.
• quality - 480p or 720p only. 480p is the budget lane within the budget tier.
• aspect_ratio - adaptive default, or the usual menu (16:9, 9:16, 1:1, 4:3, 3:4, 21:9).
• generate_audio - on by default, no extra cost. Yes, even the cheap tier includes sync audio; that's Seedance's calling card across every variant.

Outputs: video (connect a Save Video node or nothing appears), task_id, status_info, result_urls (24-hour expiry), response_json. Default timeout 30 minutes.

Install and key

The shared EvoLink routine, one last time (it's genuinely the same for every node in this pack). ComfyUI Manager → search EvoLink → install → full restart (close the console, don't just refresh the tab). Manual:

cd ComfyUI/custom_nodes
git clone https://github.com/deeplearning-goethe/comfyui-evolink

Only requests as a dependency; no models, no GPU. Key from evolink.ai/dashboard/keys (sk-), pasted once into api_key - saved locally, blank after, never in a shared workflow JSON.

Where Mini fits, and its honest limits

Use Mini as the gate at the front of a video pipeline: validate prompts, check framing, sanity-check the wiring of your generation → edit → upscale chain - all at the price point that doesn't make you wince. When a prompt proves itself, promote it to Fast or the full 2.0 node for the real render. That's the intended workflow, and it works.

The limits are equally honest: Mini is visibly the budget model, and for final, client-facing, or 4K-worthy output it's the wrong node - that's what the full Seedance 2.0 T2V is for. And no matter the tier, remember the cloud deal: every render is a line item, unchanged params cache without recharging, failed and moderation-blocked tasks don't bill. Your prompt leaves the machine. Cheap doesn't mean free - it means the meter hurts less while you're learning.