Seedance 2.0 Mini Text-to-Video API

Overview

Seedance 2.0 Mini Text-to-Video generates short video variations from text at the lowest documented iCreat rates in the Seedance 2.0 family. It is designed for high-frequency and batch workflows where teams need more hooks, localized versions, product concepts, or social variations before selecting final outputs.

The endpoint creates the scene, action, camera behavior, pacing, and optional generated audio from one required Prompt. It supports 480P and 720P output, durations from 4 to 15 seconds, automatic duration, fixed aspect ratios, and adaptive framing.

Choose Mini when generation volume and cost matter more than high-resolution delivery. Choose Seedance 2.0 Fast when lower-latency iteration is the priority, or the main Seedance 2.0 route when the selected output needs 1080P, 4K, or more demanding final-scene quality.

Use the model through:

bytedance/seedance-2-0-mini/text-to-video

Key Features

Lowest-cost Seedance generation on iCreat. Mini uses lower per-second rates than the documented Fast and mainline Seedance 2.0 routes, making it suitable for repeated tests and larger variation sets.

Batch-oriented text generation. A reusable scene brief can be adapted across products, audiences, languages, settings, hooks, or camera treatments without preparing reference assets.

480P and 720P review workflow. Use 480P for broad concept testing and 720P for selected review outputs. Mini does not provide the 1080P or 4K tiers available on the mainline route.

Optional generated audio. Set generate_audio to true to add dialogue, ambience, sound effects, or music to each generated variation.

Automatic timing and framing. Use duration: -1 when the model can choose the clip length, and use adaptive when the frame shape can follow the scene. Fixed values are better for controlled batch comparisons.

Prompt-led scene variation. The Prompt can define the subject, setting, action, camera, pacing, and sound. Change one controlled variable at a time when comparing multiple outputs.

Optional web-assisted context. The iCreat web_search tool can add external context to a request, but it does not guarantee factual or visual accuracy.

Model Comparison

Seedance 2.0 Mini vs Fast vs Mainline Text-to-Video

Seedance 2.0 Mini Text-to-Video vs HappyHorse 1.1 Text-to-Video

Inputs

This endpoint uses one required text item inside the content array. For batch work, keep the core scene structure stable and change one variable per request, such as the hook, audience, setting, language, action, or sound treatment.

Chinese Prompts can contain up to 2,000 characters, and English Prompts can contain up to 2,000 words. Long Prompts may dilute priorities, so repeated batch requests should use the same structure and concise variable changes.

Parameters

Pricing

Total Cost = Unit Price × Output Video Duration

Quick Start

This example uses a reusable vertical-ad structure. Replace the bracketed variables to create controlled variants while keeping the remaining Prompt and parameters unchanged.

curl --fail-with-body --connect-timeout 10 --max-time 60 \
  -X POST https://api.icreat.ai/v1/task/submit/bytedance/seedance-2-0-mini/text-to-video \
  -H "Authorization: Bearer ${ICREAT_API_KEY}" \
  -H "Content-Type: application/json" \
  -d '{
    "content": [
      {
        "type": "text",
        "text": "Create a 6-second vertical short-video concept for [PRODUCT]. Open with [HOOK] shown as large visual action rather than readable on-screen text. A young creator demonstrates the product in [SETTING] while the camera makes one smooth forward move. End with a clear product close-up and a confident reaction. Use bright natural lighting, fast but readable pacing, one short sound effect at the opening, light upbeat music, and no scene change."
      }
    ],
    "generate_audio": true,
    "ratio": "9:16",
    "resolution": "720p",
    "duration": 6,
    "watermark": false
  }'

The request returns a task_id.

Submit request → Receive task_id → Poll status → Retrieve result

Query the task status until it reaches SUCCEEDED, then use the same task_id to retrieve the result.

Output

Seedance 2.0 Mini Text-to-Video uses an asynchronous task flow.

The submission request returns a task_id. Task status is checked through the query-status endpoint, and the completed result is retrieved through the get-result endpoint after the task succeeds.

Detailed result fields are not listed until they are confirmed from an actual iCreat response.

Use Cases

Ad-hook testing. Generate several openings around one offer, audience problem, or product benefit before selecting the strongest direction.

Localized short-video batches. Reuse one scene structure across languages, regions, audience segments, or seasonal messages.

Catalog-scale concept clips. Produce low-cost motion concepts for many products or categories before investing in higher-resolution final assets.

Social publishing variations. Create alternate settings, actions, framing, pacing, or audio treatments for a repeated content schedule.

Storyboard and pitch options. Turn several written concepts into inexpensive motion drafts for internal review, client selection, or campaign planning.

Limitations

Mini supports 480P and 720P only. It does not provide the 1080P or 4K output tiers available on the mainline Seedance 2.0 route.

Lower-cost generation does not guarantee the same detail, complex-motion control, or final-scene quality as the mainline model. Batch results may vary, so reviewing and selecting outputs is part of the workflow.

Text-to-Video cannot lock an exact face, product design, costume, or composition without reference assets. Identity and small visual details may change across variations.

Prompts with several simultaneous actions, camera moves, scene changes, and audio instructions can weaken clarity. Keep each short variation focused on one main idea.

Readable text is not guaranteed. Logos, captions, interfaces, labels, and long written phrases may be distorted.

Generated audio can contain unclear dialogue, distortion, or timing differences. Automatic duration and adaptive ratio may also reduce consistency when comparing a controlled batch.

The provided iCreat schema does not expose a seed, so the same Prompt cannot be expected to reproduce the same video exactly. web_search may add context but does not guarantee accuracy.

FAQ

When should Seedance 2.0 Mini be chosen instead of Fast?

Choose Mini when the main goal is producing more variations at the lowest documented Seedance 2.0 family cost. Choose Fast when lower-latency iteration matters more than minimizing the cost of a larger batch.

Should a batch be generated at 480P or 720P?

Use 480P for broad testing when many concepts must be compared. Use 720P for smaller batches, selected review outputs, or cases where motion and visual detail need to be judged more clearly.

How should one Prompt be adapted into several controlled variants?

Keep the scene structure, duration, ratio, resolution, and camera direction unchanged. Change one variable at a time, such as the hook, setting, audience, language, action, or audio treatment.

How can total cost be controlled across a large batch?

Start with 480P and a short fixed duration, limit the number of variables tested at once, and move only selected concepts to 720P. Cost equals the per-second rate multiplied by the duration of every generated output.

Can the same Prompt produce consistent variations?

The same structure can produce comparable variations, but not identical videos. The endpoint does not expose a seed, so motion, framing, timing, and fine details may change between generations.