Veo API in China — Veo isn't in the catalog; these video models are
Google's Veo is not in the Router One catalog at the moment, and neither is Grok Video — the lineup follows the live catalog, so check /models for current coverage. Which video models the gateway serves, the fixed clip each one produces, and its per-clip price are read from the live catalog at /models — video models can be temporarily absent while routing is updated. The workflow does not change: submit a prompt, poll the task, download the clip — per-video unit pricing from the same wallet and key you use for chat and image models.
Direct access, no VPN
The gateway endpoint at https://api.router.one/v1 is reachable from Mainland China. No VPN, no overseas cloud console, no foreign card — an API key and a prepaid wallet are the whole setup.
Per-video unit pricing
Video models are unit-priced per generated clip — the flat USD amount is shown live in the catalog, deducted from your wallet, and recorded in the per-request cost trace. Clip length is fixed per model, so the cost of a call is known before you make it.
One key for video, image, and chat
The same key drives every video model in the catalog, the text-to-image lineup, and the full chat catalog. Per-key budgets cap a production video key independently.
Video models in the catalog
Clip length and resolution are locked per model — a duration or size field in the request body is ignored. Veo and Grok Video are not currently listed. Live per-video prices are on the models page under the video category.
| Model ID | What it is | Fixed clip |
|---|---|---|
| viduq3-turbo | Vidu Q3 Turbo, text-to-video | 5 s · 720p |
| viduq3-turbo-540p | Vidu Q3 Turbo, text-to-video, cheaper 540p tier | 4 s · 540p |
| pixverse-video | PixVerse v5, text-to-video | 5 s · 540p |
| MiniMax-Hailuo-2.3 | MiniMax Hailuo 2.3, text-to-video | 6 s · 768P |
| MiniMax-Hailuo-02 | MiniMax Hailuo 02, text-to-video | 6 s · 768P |
| wan2.6-i2v | Wan 2.6 (Alibaba Tongyi Wanxiang), image-to-video — exactly one image_url required | 5 s · 720P |
Submit, then poll
Video generation is asynchronous: the POST returns a task, and a GET on the task ID reports its status until the clip is ready. Don't wrap the call in a tight client timeout — poll instead. Full request and response fields are in the docs.
- 1Create a Router One API keySign up at router.one and create an API key in the dashboard. Top up with a card or Alipay through one hosted checkout, or with USDT/USDC on six chains (Tron, BSC, Ethereum, Polygon, Base, Arbitrum). No US credit card required.
- 2Submit a generation taskPOST to https://api.router.one/v1/videos/generations with a video model ID from the catalog and your prompt. The response returns a task ID. The image-to-video model takes an image_url as well; the text-to-video models reject one.
- 3Poll until the clip is readyGET the task by ID until it completes, then download the result. The request appears in Dashboard → Logs with its model, unit cost, and status.
# Submit a video generation task, then poll it curl -X POST https://api.router.one/v1/videos/generations \ -H "Authorization: Bearer sk-your-router-one-key" \ -H "Content-Type: application/json" \ -d '{"model": "viduq3-turbo", "prompt": "Waves lapping at the shore, dusk light on the water"}' curl https://api.router.one/v1/videos/generations/{task_id} \ -H "Authorization: Bearer sk-your-router-one-key"
No low fixed cap for normal paid usage. Abuse prevention, per-account protection limits, and upstream constraints may still apply; if a request returns 429, check Dashboard -> Logs or contact support to raise limits.
FAQ
Is Veo available?
Not at the moment — Veo is not in the catalog, and neither is Grok Video. The video models you can call today are the ones listed above; the lineup follows the live catalog, so /models is the current answer at any point in time.
Can I choose the clip length or the resolution?
No — each video model ships one fixed combination, listed in the table above, and a duration or size field in the request body is ignored. aspect_ratio is accepted by the Vidu and PixVerse models and defaults to 16:9. Pick the model whose fixed clip matches the output you need.
How is video generation billed?
Per video, at a flat unit price shown in the catalog, deducted from your metered wallet balance. Each task's unit cost is recorded in its request trace — there is no token accounting on generated video.
How long does a generation take?
It varies by model and load, which is why the API is asynchronous: submit the task, keep the task ID, and poll its status endpoint until the clip is ready rather than holding a request open.
Is this the Sora API?
No — Sora is not currently in the catalog. The video lineup today is Vidu Q3, PixVerse, MiniMax Hailuo, and Wan 2.6; check /models for current coverage as the catalog evolves.
Can the same key generate images?
Yes — the text-to-image lineup (gpt-image-2, Gemini 3 Pro Image, Gemini 3.1 Flash Image, Doubao Seedream, Grok Imagine) is served at /v1/images/generations with the same key and wallet. See the image generation API page.
Choose the next step for your video-generation workflow
This page owns the first video task. Use the paths below to select a live model, verify the asynchronous request contract, generate images instead, or fund the shared wallet.
Start generating video from China
Get your API key