Nodes/comfyui-evolink/EvoLink Kling O3 Text-to-Video (Official)
ComfyUI Node

EvoLink Kling O3 Text-to-Video (Official)

4K text-to-video with sound on tap

By deeplearning-goethe·Created 24 days ago·Updated a day ago· 0
EvoLink Kling O3 Text-to-Video (Official)
    • video
    • task_id
    • status_info
    • result_urls
    • response_json
    prompt
    duration5
    aspect_ratio16:9
    quality720p
    soundoff
    api_key
    timeout_seconds1200

    EvoLink_KlingO3T2V is the plain text-to-video member of Kling's newest generation, and it's the least fussy node in the O3 family: prompt in, up to 4K clip out, no reference images required. Kling's O3 line is the current flagship from Kuaishou, and this node carries the full feature set - up to 4K, generated sound, any duration from 3 to 15 seconds. If you want a hero-quality clip straight from a description, this is the Kling node that does it without any extra inputs to manage.

    The inputs that matter

    • prompt (required) - shot and motion description, up to 2500 characters, Chinese-friendly. The tooltip points at multi-shot setups in the API docs if you want to go advanced.
    • duration - 3–15 seconds, default 5, billed by duration.
    • quality - 720p (default), 1080p, or 4k.
    • aspect_ratio - 16:9 (default), 9:16, or 1:1. Vertical shorts → 9:16.
    • sound - off (default) or on. Generated sound effects; this is where the O3 tier beats Kling Turbo, which doesn't support sound at all.

    Then the pack-standard api_key and timeout_seconds (1200).

    How it works

    The job posts to POST /v1/videos/generations as kling-o3-text-to-video, and the pack polls the task until the clip completes, showing a progress bar on the node. The finished video downloads to your machine as a VIDEO object - wire video into a native Save Video node to keep it. The other outputs are task_id (dashboard logs), status_info (model/runtime/credits), result_urls (24h expiry), and response_json. Failed or blocked jobs don't charge, and rerunning unchanged params reuses ComfyUI's cache.

    Because the output is a VIDEO type, you need a recent ComfyUI with comfy_api support - old installs error with a message about missing VIDEO output.

    Installing

    No GPU, no model files, only requests:

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

    then fully restart ComfyUI, or install via ComfyUI Manager by searching EvoLink.

    Troubleshooting

    Pack-standard errors (401 key, 402 balance, 403 access, 429 rate limit) and the VPN note - whitelist api.evolink.ai and files-api.evolink.ai. Budget-wise, remember 4K costs a lot more than 720p per run; use the cheap tier for drafts and keep 4K for the keepers. And if you enable sound, expect the run to take longer - audio generation rides along with the video pass, and the timeout default of 1200s exists for a reason.

    CategoryEvoLink

    Inputs (7)

    NameTypeDefaultDescription
    promptSTRING镜头/画面描述,支持中文,最长 2500 字符。多分镜等高级用法见 evolink.ai/docs。
    durationoptINT53–15视频时长(秒),3~15 任意整数,按时长计费
    aspect_ratiooptCOMBO16:9画幅比例
    qualityoptCOMBO720p分辨率
    soundoptCOMBOoff生成音效
    api_keyoptSTRINGEvoLink API Key(sk- 开头)。首次填写后自动保存到本机配置,之后可留空。分享工作流前请清空此框。获取:evolink.ai/dashboard/keys
    timeout_secondsoptINT120060–3600最长等待时间;生成失败或审核拦截的任务不扣费

    Outputs (5)

    NameTypeDescription
    videoVIDEO生成结果(已下载落地,可直连保存节点)
    task_idSTRINGEvoLink 任务 ID,可到 evolink.ai/zh/dashboard/logs 查询
    status_infoSTRING人读状态摘要(模型/用时/消耗 credits)
    result_urlsSTRING结果原始链接,每行一个(24 小时过期)
    response_jsonSTRING平台 GET /v1/tasks/{id} 的完整 JSON 响应(status/usage/results 等),供下游节点解析