grok / v1.5 / textToVideo

The v1.5 text-to-video node is what you reach for when you want Grok's closed model but you don't have a start image to work from - and you want the option of 1080p, which the v1.0 text node never had. Prompt in, clip out, Buzz charged. It's the simplest node in the Grok v1.5 set: exactly one required input, no image sockets, no reference tagging, no JSON. Just prompt engineering and a cost meter.

It's part of civitai-comfy-nodes, Civitai's official ~160-node pack generated from its Orchestration API spec. The node submits a videoGen job tagged grok / v1.5 / textToVideo to Civitai's cloud, long-polls until it finishes, and returns a native VIDEO tensor. Nothing runs on your GPU - the model is xAI's closed video stack, and the only door into a Comfy graph is an API wrapper like this one.

What you set

• prompt (required, multiline) - the whole job. There's no image conditioning to lean on, so prompt quality is the entire game: subject, camera movement, lighting, scene.
• duration - 1 to 15 seconds, default 6.
• resolution - 480p, 720p, or 1080p (v1.5's headline addition over v1.0).
• aspect_ratio - no auto; pick 16:9 (default), 4:3, 3:2, 1:1, 2:3, 3:4, 9:16.

Outputs: video (VIDEO), draft_cache_audio (AUDIO), draft_cache_video (VIDEO), workflow_id and raw_json (STRING). The node shows its workflow id and cost on itself after each run.

Install and auth

Same as the whole pack - official, Comfy Registry:

# ComfyUI Manager: search "Civitai Comfy Nodes"
cd ComfyUI/custom_nodes
git clone https://github.com/civitai/civitai-comfy-nodes.git
pip install -r civitai-comfy-nodes/requirements.txt

Restart and authenticate via a Civitai Auth node, the CIVITAI_API_TOKEN env var, or the sidebar OAuth login.

Common issues

1080p is where the bill grows. Text-only generation has no image to keep it grounded, so you'll retry for composition - and each retry at 1080p is the expensive kind. Draft at 480p/720p and promote only the winner. Prompt discipline pays double here. With no start frame, vague prompts produce wandering video; get the camera and subject into the first sentence. Your prompt leaves your machine - it's sent to a third-party server, and that's the deal with every API wrapper. Early preview applies: the pack README warns nodes and behavior can change without notice. And if you find yourself wanting a start image after all, the v1.5 imageToVideo and referenceToVideo nodes sit right next to this one in the same Civitai/Video/grok menu - same engine, more conditioning, more control.