# Poyo Grok Imagine

> Grok Imagine video generation on PoYo / poyo.ai via `https://api.poyo.ai/api/generate/submit`; use for `grok-imagine`, text-to-video, image-to-video, 6s/10s clips, and `fun` / `normal` / `spicy` mode control.

- Skill: `dvcrn/poyo-grok-imagine-2` (Agent Skill, multi-file: 3 files)
- Install (CLI): `npx skillmds@latest add dvcrn/poyo-grok-imagine-2`
- Raw SKILL.md: https://api.skillmd.com/api/skills/dvcrn/poyo-grok-imagine-2/raw
- Safety review: pending (external: skill-scanner PASS, skillspector PASS)
- Works with: Claude Code, Claude.ai, OpenAI Codex
- Category: Integrations & APIs
- Author: dvcrn (https://skillmd.com/u/dvcrn)
- Updated: 2026-09-09
- Page: https://skillmd.com/skills/dvcrn/poyo-grok-imagine-2

---


# PoYo Grok Imagine Video Generation

Use this skill for `grok-imagine` jobs on PoYo. It covers short text-to-video, image-to-video, and mode-based styling.

## Use When

- The user explicitly asks for `Grok Imagine` or `grok-imagine`.
- The task is a 6-second or 10-second clip.
- The workflow needs text-to-video, image-to-video, or `mode` styling.

## Core Capability

- `grok-imagine` is a single-model video entry point. Use `image_urls` for image-to-video and `mode` for `fun`, `normal`, or `spicy` style control.

## Key Inputs

- `prompt` is required.
- `image_urls` is for image-to-video and supports one image.
- `duration` supports `6` and `10`.
- `aspect_ratio` supports `1:1`, `2:3`, `3:2` for text-to-video.
- `mode` supports `fun`, `normal`, `spicy`.

## Execution

- Read `references/api.md` for endpoint details, model ids, key fields, example payloads, and polling notes.
- Use `scripts/submit_grok_imagine.sh` to submit a raw JSON payload from the shell.
- If the user only needs a curl example, adapt the example from `references/api.md` instead of rewriting from scratch.
- After submission, report the `task_id` clearly so follow-up polling is easy.

## Output expectations

When helping with this model family, include:
- chosen model id
- final payload or a concise parameter summary
- whether reference images are involved
- returned `task_id` if a request was actually submitted
- next step: poll status or wait for webhook