Cheapest Seedance 2.5 API Access: Compare Real Costs
Protoface has the lowest standard Seedance 2.5 rates in this comparison. Compare Flash, Sume’s provider-price access, billing minimums and failure policies.

For standard Seedance 2.5 without video input, Protoface has the lowest published rates among the options compared here, starting at $0.084 per output second. Sume starts at $0.1028 per second with provider-price billing. If a different model tier fits your workload, OpenRouter lists Flash from $0.06726 per second, before credit-purchase fees.

This comparison covers standard-model access through Protoface, BytePlus ModelArk, Sume and Runway, with Flash considered separately. The criteria are resolution-matched generation rates, billing minimums, funding costs and failed-generation policies. Public prices and terms were checked on November 18, 2026. The lowest rate in this set is not a market-wide lowest-price guarantee.
What is the cheapest way to use Seedance 2.5 through an API?
Start by deciding whether you need standard Seedance 2.5 or Flash. OpenRouter describes Flash as a separate, lower-cost tier with output at 480p and 720p. Do not treat its starting price as a cheaper quote for the identical standard model. See OpenRouter’s Flash model listing.
For standard text-to-video, compare the rates below at the same resolution, duration, aspect ratio and audio setting. Then apply each API’s rounding and minimum-charge rules.
Sume’s cost-control workflow is straightforward: estimate the exact input before generation, then submit it. The estimate and charge use the same pricing function, as documented in the Sume API reference.
Seedance 2.5 API pricing by resolution
The table covers standard text-to-video without input or reference video. Rates are USD per output second, checked November 18, 2026. They exclude funding fees and taxes, negotiated contracts and introductory credits.
| API access | Billing basis | 480p | 720p | 1080p |
|---|---|---|---|---|
| Protoface | Output seconds, then round each request up to a cent | $0.084 | $0.184 | $0.464 |
| BytePlus ModelArk | Actual generation tokens; figures shown are estimates | ≈$0.1028 | ≈$0.23112 | ≈$0.56862 |
| Sume | Published output-second rates | $0.1028 | $0.23112 | $0.56862 |
| Runway | Output seconds paid in credits | $0.20 | $0.30 | $0.68 |
BytePlus bills tokens rather than a flat per-second tariff. Its estimates above apply its published formula to widescreen output at 24 fps: 854 × 480, 1280 × 720 and 1920 × 1080 pixels. Actual settlement uses usage.completion_tokens. BytePlus’s rounded examples are consistent with these estimates. See BytePlus’s token rates and billing formula.
Flash pricing is a separate comparison
BytePlus lists Flash without video input at $7 per million generation tokens, versus $10.70 for standard Seedance 2.5 at the lower resolutions. Its estimated five-second Flash examples cost approximately $0.34 at 480p and $0.76 at 720p. See BytePlus’s Flash pricing.
OpenRouter’s starting Flash rate is an inference price, not the complete cost of funding an account. Its ordinary credit-purchase fee is 5.5%, with a $0.80 minimum; crypto payments have a separate 5% fee. Sources: OpenRouter Flash pricing and OpenRouter’s fee schedule.
Keep standard generation separate from Remix
Sume’s Remix Motion Transfer, Object Swap and Restyle listings charge per source second. The standard Seedance rates above are per output second. A Remix quote is not a quote for the same text-to-video request. Both billing units appear on Sume’s API pricing page.
Does zero markup mean the lowest total API cost?
No. Zero markup describes the relationship between a platform’s rate and its provider cost. It does not establish that every other platform charges more.
Sume offers provider-price access without a subscription or an added card fee. Funding starts at $5, and each top-up lasts 365 days. See Sume’s pricing and funding terms.
OpenRouter also states that it passes through provider inference prices, but charges separately when you purchase credits. Its minimum fee makes small purchases proportionally more expensive than the headline percentage suggests. See OpenRouter’s pricing and fees.
Compare in this order:
- Model: standard or Flash.
- Request: resolution, aspect ratio, duration, audio and reference media.
- Purchase terms: public usage rates, prepaid credits, subscriptions or negotiated pricing.
- Payable charge: request minimums, rounding and funding fees.
A difference between public listings does not, by itself, establish a reseller’s margin. A markup calculation needs the underlying provider cost for the same configuration and purchase terms. The table therefore compares published rates rather than assigning inferred markup percentages.
How do failed generations and minimum billing affect cost?
Failure refunds are not the same as creative retries
| API | Documented failure or cancellation treatment |
|---|---|
| Sume | Failed and canceled jobs are refunded. Cancellation is available until generation starts. |
| BytePlus | Charges only for successfully generated videos, with content-moderation failures included among failures that incur no fee. |
| Protoface | Failed runs release reserved credits. Cancellation before a run starts releases the reservation; cancellation after it starts charges the full reserved amount. |
| Runway | Its Seedance guide states that tasks failing input moderation are still charged. |
Keep API failures separate from completed takes your team rejects. A refund policy for failed jobs does not answer how much you will spend generating technically completed alternatives.
For budgeting, use clearly defined metrics:
Delivery cost per second =
net generation spend after refunds / successfully delivered output seconds
Production cost per accepted second =
net generation spend after refunds / output seconds accepted for use
Add funding fees explicitly if you want a fully loaded metric. Use the same definition for every API. Neither metric predicts acceptance rates without results from your own workload.
Minimum duration and rounding change short-clip costs
Protoface bills at least four output seconds and rounds each request’s total up to the next cent. A five-second request costs $0.42 at 480p or $0.92 at 720p, calculated from Protoface’s rates and rounding rules.
Runway’s standard Seedance minimum is 80 credits per generation. At $0.01 per credit, that is an $0.80 minimum before any additional video-reference charges. See Runway’s API pricing.
Sume accepts durations of 4-30 seconds and rounds each request up to $0.0001. Use the exact-input estimate rather than rounding the displayed rate yourself. See Sume’s duration and pricing rules.
Video-input workloads need a separate calculation. BytePlus includes input-video duration and applies minimum token consumption. Protoface uses replacement input-plus-output rates with additional duration rules. Do not extend the text-to-video table to editing, extension or reference-video requests. Sources: BytePlus video billing and Protoface video-input billing.
Account funding is not a generation charge
Sume’s $5 funding minimum is a balance purchase, not a charge for your first video. See Sume’s funding terms.
BytePlus’s activation requirements specify a qualifying available balance of $30 under Pay by Credits, a qualifying dedicated AI Savings Plan of at least $30, or a qualifying resource pack of at least $30. Eligibility, validity and scope differ by route. See BytePlus’s activation requirements.
Separate the cash needed to activate an account from the amount consumed by generation. Before depositing funds, check checkout minimums, fees and balance-expiration terms.
How much would your Seedance 2.5 workload cost?
Consider a hypothetical batch of 100 completed five-second standard videos at 720p, 16:9, without video references. Use the same audio setting across APIs. For the BytePlus estimate, assume 1280 × 720 output at 24 fps.
| API | Calculation from published pricing | Generation spend |
|---|---|---|
| Protoface | 100 × $0.92 per request | $92.00 |
| BytePlus | Estimated 500 output seconds × $0.23112 | ≈$115.56 |
| Sume | 500 output seconds × $0.23112 | $115.56 |
| Runway | 500 output seconds × $0.30 | $150.00 |
The calculations apply the linked rates checked November 18, 2026. Protoface’s cent rounding and Runway’s minimum do not increase this batch: each Protoface request totals an exact cent amount, and each Runway request exceeds its minimum. BytePlus remains an estimate because actual tokens determine settlement.
These totals exclude failed attempts, extra completed takes, taxes and funding fees. They do not compare measured quality, latency or first-pass acceptance. Flash is excluded because switching tiers changes the model being compared. An unused prepaid balance is not generation spend consumed by this batch.
How to start using Seedance 2.5 through Sume
Create an API key and store it server-side as SUME_API_KEY. The model identifier is seedance-2.5. The following request structure comes from the Sume API reference.
Estimate the exact request first
This call estimates a text-only request without generating a video:
curl -X POST https://api.sume.com/v1/open/estimate \
-H "Content-Type: application/json" \
-d '{
"model": "seedance-2.5",
"input": {
"prompt": "A slow camera move around a ceramic lamp in a sunlit studio.",
"duration": 5,
"resolution": "480p",
"aspect_ratio": "16:9",
"generate_audio": false
}
}'
Submit the same input with server-side authentication
Send model and the fields inside input as one top-level JSON object to POST https://api.sume.com/v1/video-router/generate, with Authorization: Bearer $SUME_API_KEY. Save the job ID, poll GET /v1/jobs/{jobId}/status, and retrieve completed output from GET /v1/jobs/{jobId}/result. See Sume’s submission and queue instructions.
Before scaling, record a representative request’s settings, estimate, actual charge and acceptance decision. Start with Sume’s API pricing, then estimate the exact request your application will send.