Short answer: Seedance 1.5 Pro is ByteDance Seed’s joint audio-video generation model. It is worth evaluating if synchronized sound, dialogue and lip-sync, or camera direction are central to your video workflow. But don’t assume that a current, production-ready API route for version 1.5 is available just because the model page has a “Get API” button. The Volcano Engine video API documentation reviewed here centers on newer Seedance versions; confirm that your chosen provider currently enables 1.5, and verify its model ID, terms, and pricing before you build around it.
What is Seedance 1.5 Pro?
ByteDance describes Seedance 1.5 Pro as a joint audio-video model that follows complex instructions. Its product page claims coordinated voice and spatial sound generation, support for a range of languages and dialects, lip-sync and motion alignment, complex camera movement, and narrative cohesion. The page names short dramas, advertising, and social media as potential use cases. These are publisher claims—not guarantees for a particular prompt, language, or production workflow. (ByteDance’s Seedance 1.5 Pro overview)
The technical-report abstract describes the model as a foundation model for native audio-video generation and claims its inference acceleration framework delivers a speed increase of more than 10×. The abstract does not specify the baseline or measurement conditions for that figure, so treat it as a paper-reported claim rather than a throughput estimate for your own service. (Seedance 1.5 Pro technical report abstract)
The official page refers to internal SeedVideoBench-1.5 results, but the available page text does not provide scores or enough methodology to establish how the model compares with alternatives. Don’t use that reference alone as evidence that it will outperform another model on your task. (ByteDance’s internal benchmark overview)
When is it a good candidate?
Put Seedance 1.5 Pro on a shortlist when your output needs generated video and coordinated audio together—for example, dialogue-led social clips, short narrative scenes, or advertising concepts where camera movement and expressive performance matter. That’s a reason to test it, not proof that it will meet a production brief.
Be more cautious if your delivery depends on a particular dialect, precise lip-sync, repeatable camera instructions, or continuity across several clips. The official page makes relevant qualitative claims, but the material reviewed does not establish measured performance for your particular language, content, or acceptance criteria. Likewise, if you need exact resolution, frame rate, duration, or camera-control guarantees, obtain those specifications from the provider serving your request before committing.
A practical evaluation is to run the same small set of representative prompts through Seedance and any alternatives you are considering. Include at least one dialogue scene in the target language, one prompt with a required camera move, and one multi-shot or recurring-character task if that matters to your project. Score the actual outputs against your own requirements: intelligible audio, acceptable lip-sync, followed directions, continuity, usable exports, and the share of attempts that pass. This is a proposed evaluation—not a test performed for this article.
Can you use Seedance 1.5 Pro through a production API?
The public material reviewed here does not establish a currently supported, production-ready API path for Seedance 1.5 Pro. ByteDance’s model page displays a “Get API” call to action, and the technical-report abstract says the model is accessible on Volcano Engine. Neither statement, by itself, supplies working integration instructions or confirms current access for your account. (ByteDance model page; technical-report abstract)
Volcano Engine’s video-generation tutorial describes an asynchronous workflow: submit a task, receive a task ID, then poll for completion or use a webhook. However, the example shown uses a Seedance 2.5 model ID. The video-generation model overview reviewed here lists Seedance 2.5, 2.0, and 1.0 Pro—not a 1.5-specific integration. Treat these as evidence of the platform’s general video-task pattern and the versions listed in that overview, not confirmation that the same route currently accepts Seedance 1.5 Pro.
Before integrating, check the chosen provider’s current documentation or account console for:
- Whether Seedance 1.5 Pro specifically is available to your account and in your intended region.
- The exact model or endpoint identifier and version.
- Authentication, request fields, accepted input formats, and response structure.
- Whether jobs run asynchronously, and the supported polling or webhook process.
- Generation limits, failure behavior, output retention, and relevant data-handling terms.
- Service-level or reliability terms that apply to your deployment.
If the provider only confirms Seedance 2.x or another version, don’t silently treat it as 1.5. Version, capabilities, and commercial terms may differ.
What does Seedance 1.5 Pro cost?
There is no single price established by the model description; cost depends on the provider and access channel. The Coze model-cost page lists Seedance 1.5 Pro audio-video generation at 16 points or ¥0.016 per 1,000 input-plus-output tokens, and silent-video generation at 8 points or ¥0.008 per 1,000 input-plus-output tokens. The page identifies those as Coze model rates; it says Coze costs default to point deduction, with cash taken from the main account if the point balance is insufficient. These figures are for the Coze-hosted model—not confirmed direct Volcano Ark API rates. The retrieved pricing page does not show an effective date, and the cost of a typical finished clip still depends on its billable token usage.
For a production estimate, use the price and billing unit displayed by the provider you will actually call. Confirm that the quoted rate applies to your API route, then estimate spend from expected successful outputs, tests, and retries. Don’t substitute free credits from a hosted web generator or an unattributed “cheaper than official” claim for API pricing.
Bottom line
Evaluate Seedance 1.5 Pro if native audio-video generation and the model’s claimed dialogue, lip-sync, camera, or narrative capabilities match your use case. Do not treat model selection as API approval: confirm 1.5-specific availability and integration details with the intended provider, and calculate costs from that provider’s applicable price and billing unit. If either the API route or commercial terms remain unconfirmed, keep the model on a shortlist rather than making it a production dependency.
Sources
Seedance 1.5 pro — ByteDance Seed
Seedance 1.5 pro: A Native Audio-Visual Joint Generation Foundation Model
Video Generation Tutorial — Volcano Engine