Which ByteDance models can I call through Felo?

Four Seedance video routes: Seedance 2.5, Seedance 2.0, Seedance 2.0 Fast and Seedance 2.0 Mini. ByteDance's Seedream image models are catalogued under the separate ByteDance Seed provider entry.

Are Seedance models text-to-video only?

No. They accept text, image, video and audio references and return a video with synchronized audio; the exact input modes available depend on the tier.

How are Seedance routes priced on Felo?

Per output video token in USD, with rates that vary by input mode and resolution. Reference-video requests are cheaper than text-to-video, and audio-on and audio-off tiers carry the same rate.

What separates Seedance 2.5 from Seedance 2.0?

Seedance 2.5 extends output to 30 seconds in one pass with two further extension rounds and up to 50 reference inputs in total, while Seedance 2.0 generates 15-second multi-shot clips from up to 12 input assets.

Do Seedance outputs carry a watermark?

Seedance 2.0 applies C2PA watermarking to every video, and Seedance 2.5 keeps watermark and IP guardrails on output.

Do I need a ByteDance account to use Seedance on Felo?

No. The Seedance routes on this page are called with a Felo API key; a ByteDance cloud or BytePlus account is only needed to use the models from ByteDance directly.

ByteDance API

ByteDance's Seed team builds the Seedance family, a line of audio-video generation models that take text, images, video and audio references and return a video with synchronized sound. Felo carries four Seedance routes: 2.5, 2.0, 2.0 Fast and 2.0 Mini, all called with a Felo API key rather than a ByteDance cloud account.

Official website

Models: 5; image ×1 · video ×4
Input per 1M: Not published
Output per 1M: Not published
Largest context: —

What ByteDance Seed publishes

ByteDance Seed was established in 2023 and works across language models, speech, vision, world models and AI infrastructure. Its public portfolio includes the Seed2.1 agent model, Seedance for video, Seedream for image generation, SeedRealtime and Seed Audio 1.0.

Felo exposes the Seedance video line on this page. ByteDance's Seedream image models appear in the catalog under the separate ByteDance Seed provider entry.

The Seedance line

Seedance 2.5 generates up to 30 seconds of video in one pass and supports two further extension rounds, with up to 30 image, 10 video and 10 audio references and timestamp- or region-level editing. Seedance 2.0 is the 15-second multi-shot model with native synchronized audio and C2PA watermarking on output, while 2.0 Fast and 2.0 Mini are the speed- and cost-optimized tiers.

How Seedance is billed on Felo

The Seedance routes are billed per output video token in USD, with different rates by input mode and resolution. Reference-video requests bill at a lower rate than text-to-video, and the catalog's rates are identical whether audio generation is on or off.

The provider table's per-token input and output columns do not apply to these video routes; the catalog's per-route video rates are the authoritative figures for the current price.

Fit, access and limits

Seedance is aimed at multi-shot narrative film, marketing and social campaigns, e-commerce showcases and short-form social content, with the Fast and Mini tiers used for drafts, volume and iteration. Access runs through ByteDance's own platforms and third-party providers, and the Seedance site links to the BytePlus playground for trying the models; on Felo the same API key calls these routes.

Output length depends on the tier: up to 30 seconds for 2.5, 15 seconds for 2.0, and 4-15 seconds for 2.0 Fast.

Doubao-Seedream-5.5

Input per 1M: Not published
Output per 1M: Not published
Context: —
Max output: —

Seedance 2.0 Mini

Input per 1M: Not published
Output per 1M: Not published
Context: —
Max output: —

Seedance 2.5

Input per 1M: Not published
Output per 1M: Not published
Context: —
Max output: —

Seedance 2.0

Input per 1M: Not published
Output per 1M: Not published
Context: —
Max output: —

Seedance 2.0 Fast

Input per 1M: Not published
Output per 1M: Not published
Context: —
Max output: —

ByteDance models compared by input price, output price, context window and maximum output.
Model	Input per 1M	Output per 1M	Context	Max output
Doubao-Seedream-5.5	Not published	Not published	—	—
Seedance 2.0 Mini	Not published	Not published	—	—
Seedance 2.5	Not published	Not published	—	—
Seedance 2.0	Not published	Not published	—	—
Seedance 2.0 Fast	Not published	Not published	—	—

Prices are published standard rates in USD per 1M tokens, before channel discounts.

Frequently asked questions

Which ByteDance models can I call through Felo?: Four Seedance video routes: Seedance 2.5, Seedance 2.0, Seedance 2.0 Fast and Seedance 2.0 Mini. ByteDance's Seedream image models are catalogued under the separate ByteDance Seed provider entry.
Are Seedance models text-to-video only?: No. They accept text, image, video and audio references and return a video with synchronized audio; the exact input modes available depend on the tier.
How are Seedance routes priced on Felo?: Per output video token in USD, with rates that vary by input mode and resolution. Reference-video requests are cheaper than text-to-video, and audio-on and audio-off tiers carry the same rate.
What separates Seedance 2.5 from Seedance 2.0?: Seedance 2.5 extends output to 30 seconds in one pass with two further extension rounds and up to 50 reference inputs in total, while Seedance 2.0 generates 15-second multi-shot clips from up to 12 input assets.
Do Seedance outputs carry a watermark?: Seedance 2.0 applies C2PA watermarking to every video, and Seedance 2.5 keeps watermark and IP guardrails on output.
Do I need a ByteDance account to use Seedance on Felo?: No. The Seedance routes on this page are called with a Felo API key; a ByteDance cloud or BytePlus account is only needed to use the models from ByteDance directly.