TheRouter.aiTheRouter.ai
Back to Models

Vidu Q1 Start-End

zhipuzhipu/viduq1-start-end

Zhipu AI Vidu Q1 Start-End β€” 5s 1920Γ—1080 video interpolating between first and last frame.

Vidu Q1 model guide β€” Zhipu AI official docs β†—Video generation async API β€” Zhipu AI official docs β†—

Vidu Q1 Start-End is Zhipu AI's first-and-last-frame video generation model. It takes two input images, treats the first as the opening frame and the second as the ending frame, then generates the motion between them as a fixed 5-second, 24 fps, 1080p video. In the Vidu Q1 family, this is the controlled-transition option: use it when the visual start and destination both matter more than free-form text-to-video exploration.

The underlying Q1 release focuses on high-quality video creation: sharper image texture, fewer hand-intersection artifacts, reduced frame shake, stronger realistic scenes, better 2D animation style retention, and smoother start-end transitions. Zhipu positions it for film, advertising, animation shorts, short-drama production, and cultural-tourism promotion. Through TheRouter, use the public model id zhipu/viduq1-start-end; operational pricing and routed parameters come from the model catalog, while the official upstream API names the model viduq1-start-end.

Best for
  • β€’ Planned transitions between two known frames, such as product before-and-after shots, scene changes, or ending-card reveals
  • β€’ Advertising and e-commerce motion where a supplied product image must remain anchored at the beginning or end of the clip
  • β€’ Animation and short-drama storyboarding that needs a smooth, director-controlled 5-second bridge between keyframes
  • β€’ Cinematic or cultural-tourism clips where the final composition is known in advance and free-form drift would be costly
Reach for something else if
  • β€’ Pure text-to-video prompts with no visual anchors β€” route to zhipu/viduq1-text instead
  • β€’ Single-image animation where only the first frame is available β€” route to zhipu/viduq1-image
  • β€’ Long clips or variable-duration timelines β€” Vidu Q1 Start-End is documented as fixed 5-second output
  • β€’ Tasks requiring verified snippets today β€” code examples remain unproven until a human approves a funded snippet verification run

How TheRouter serves this differently from the vendor

As the vendor operates it

Zhipu's native MaaS API exposes Vidu Q1 Start-End at /api/paas/v4/videos/generations using upstream model name viduq1-start-end and accepts image_url as an array of two images.

On TheRouter

TheRouter serves the same start-end capability under public id zhipu/viduq1-start-end through its routed video-generation surface with unified TheRouter auth, pricing, and parameter catalog.

docs.bigmodel.cn β†—
Context Length
--
Max Output
--
Request Priceper request
$0.432/ request

Modalities

imagetext→video

Capabilities

VisionImage to Video

Media Generation Capabilities

video_generation
sizes
  • 1920x1080
durations_seconds
  • 5
defaults
size
1920x1080
duration_seconds
5

Pricing Breakdown

TypeRate
Request$0.432 / request

Supported Parameters

promptimage_urldurationsizemovement_amplitude

Specifications

Generation modeStart/end frame to videodocs.bigmodel.cn β†—verified
Input imagesTwo images: first image is used as the starting frame; second image is used as the ending framedocs.bigmodel.cn β†—verified
Duration5 secondsdocs.bigmodel.cn β†—verified
Frame rate24 fpsdocs.bigmodel.cn β†—verified
Output resolution1080p / 1920Γ—1080docs.bigmodel.cn β†—verified
Upstream model nameviduq1-start-enddocs.bigmodel.cn β†—verified
Upstream list priceΒ₯2.5 per requestdocs.bigmodel.cn β†—verified
Generated URL validity1 daydocs.bigmodel.cn β†—verified
Concurrent in-flight tasksV0: 5, V1: 10, V2: 15, V3: 20docs.bigmodel.cn β†—verified

Benchmarks

BenchmarkDistributionScoreSource
Official qualitative quality claims
Zhipu describes stronger clarity, fewer hand-intersection artifacts, reduced frame shake, near-realistic style, better 2D animation style retention, and smoother start-end transitions, but does not publish a numeric benchmark table for Vidu Q1 Start-End on this page.
Not numerically discloseddocs.bigmodel.cn β†—
Start-end transition quality
The official guide says Q1's start-end transition is smoother and highlights dynamic-frame prediction plus style-fusion methods. No numeric score is disclosed.
Qualitatively improveddocs.bigmodel.cn β†—
Output specification
All Vidu Q1 modes are documented as fixed 5-second, 24 fps, 1080p output, with the start-end request example using duration 5 and size 1920x1080.
5s / 24 fps / 1080pdocs.bigmodel.cn β†—

API Usage Examples

Use the global api.therouter.ai endpoint shown below for new integrations; the legacy China accelerated endpoint is retired.

cURL
curl https://api.therouter.ai/v1/videos   -H "Content-Type: application/json"   -H "Authorization: Bearer $THE_ROUTER_API_KEY"   -d '{
    "model": "zhipu/viduq1-start-end",
    "prompt": "A slow tracking shot through a neon-lit city street",
    "duration": 8
  }'

API guide

Full API reference β†’

Async video generation

Submit a first frame, a last frame, and a prompt to create a fixed 5-second 1080p transition video. Snippets are illustrative only until human-approved verification is funded.

cURL
curl https://api.therouter.ai/v1/videos \
  -H "Authorization: Bearer $THEROUTER_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "zhipu/viduq1-start-end",
    "prompt": "A plush rabbit drives from a quiet street into a bright spring market, smooth cinematic transition",
    "image_url": [
      "https://example.com/first-frame.jpg",
      "https://example.com/last-frame.jpg"
    ],
    "duration": 5,
    "size": "1920x1080",
    "movement_amplitude": "auto"
  }'

More from zhipu

zhipu/viduq1-imageUse when only the first frame image is supplied and the model may invent the endingzhipu/viduq1-textText-to-video sibling for prompts without image anchorszhipu/vidu2-start-endLower-cost Vidu 2 start-end sibling for 4-second 720p transitions

Similar models

Cross-provider sibling models
doubao/doubao-seedance-2-0
Higher-end video model when you need broader multimodal controls and native audio
doubao/doubao-seedance-2-0-fast
Fast video-drafting alternative for lower-cost iterative storyboard tests

News & changes

2025-06-18

Zhipu documents Vidu Q1 for 5-second 1080p video generation

The official Vidu Q1 guide describes image-to-video, start-end-frame, and text-to-video modes, all fixed at 5 seconds, 24 fps, and 1080p. The start-end mode accepts two images and uses them as the first and last frame for smoother controlled transitions.

re-authored by TheRouterdocs.bigmodel.cn β†—

Frequently asked

What does Vidu Q1 Start-End use the two images for?

The first image is treated as the starting frame and the second image is treated as the ending frame. The model generates the video between those frames, so it is best for transitions where both endpoints are already designed.

re-authored by TheRouterdocs.bigmodel.cn β†—
How long and what resolution are the generated videos?

The official Vidu Q1 page lists Start-End output as 5 seconds at 1080p, with the broader overview specifying 24 fps. The upstream call example uses duration 5 and size 1920x1080.

re-authored by TheRouterdocs.bigmodel.cn β†—
Are the snippets on this page verified against the live API?

No. This scheduled curation run has no paid API budget, and snippet verification requires an operator-set approval flag. Treat the snippets as integration-shaped examples until a human funds and approves verification.

re-authored by TheRouterdocs.bigmodel.cn β†—
Fact ledger β€” every claim on this page traces here
sourceURLretrieved
Generation modedocs.bigmodel.cn β†—2026-08-18verified
Input imagesdocs.bigmodel.cn β†—2026-08-18verified
Durationdocs.bigmodel.cn β†—2026-08-18verified
Frame ratedocs.bigmodel.cn β†—2026-08-18verified
Output resolutiondocs.bigmodel.cn β†—2026-08-18verified
Upstream model namedocs.bigmodel.cn β†—2026-08-18verified
Upstream list pricedocs.bigmodel.cn β†—2026-08-18verified
Generated URL validitydocs.bigmodel.cn β†—2026-08-18verified
Concurrent in-flight tasksdocs.bigmodel.cn β†—2026-08-18verified
Official qualitative quality claimsdocs.bigmodel.cn β†—2026-08-18verified
Start-end transition qualitydocs.bigmodel.cn β†—2026-08-18verified
Output specificationdocs.bigmodel.cn β†—2026-08-18verified
Zhipu documents Vidu Q1 for 5-second 1080p video generationdocs.bigmodel.cn β†—2026-08-18verified
What does Vidu Q1 Start-End use the two images for?docs.bigmodel.cn β†—2026-08-18to verify
How long and what resolution are the generated videos?docs.bigmodel.cn β†—2026-08-18to verify
Are the snippets on this page verified against the live API?docs.bigmodel.cn β†—2026-08-18to verify
Help & contact