Try Gemini Omni Flash 1.2 Video Upscale from Google →
Models
Agents
Workflows
Studio
PricingBlogDocs
ExploreDiscover models by categoryDiscoverModels and effects built by WiroBrowse All ModelsBrowse the complete catalogSee FavoritesSign in to view saved models
Generative Media AgentCreate and edit media by chattingWorkflow AgentBuild visual workflows with Agent
OverviewThe platform at a glanceLearnSkills, knowledge, guardrailsAnatomyWhat makes agents reasonBuild Your AgentPick connectors and skills, choose a plan, deploy
Pre-built AgentsBrowse the catalog
Agent Usecases
Ad Campaign ManagerApp Event ManagerApp Review RepliesBarber BookingCustomer Win-BackEcommerce ListingsRestaurant Reviews
Sign InStart Building
Explore models/
Video GenerationActive

Seedance v2.5 Flash Reference Uncensored

bybytedance

ByteDance Seedance v2.5 Flash generates or edits 4–30s videos from reference images, clips, and audio. This uncensored variant can also generate synced audio.

Text to VideoImage to Video
Model ID
seedance v2.5 flash reference uncensored
Provider
bytedance
Updated
Nov 16, 2026, 11:40
0
Comments
Average rating : 3 (2 users)
Providerbytedance
Modelseedance v2.5 flash reference uncensored
Text to VideoImage to Video
wiro playground—bytedance/seedance v2.5 flash reference uncensored
1 / 10
Maximum 10 image allowed
Delete All
Image 1

Optional. The provider may sometimes reject human faces. 1-10 reference images. In the prompt, refer to them as [Image 1], [Image 2], etc. Reference images are not charged.

Optional. The prompt to generate the video. Refer to reference videos as [Video 1], images as [Image 1], audios as [Audio 1], and say what to keep and what to change.

Required. Video resolution.

Required. Generated video duration in seconds. Use 'auto' to let the model choose the length. 'auto' is required when you edit a reference video (changing its lighting, style or objects), because the output then keeps the length of the input video. A number is only used when you generate from references or extend a video, where it sets the length of the new footage. The model determines the task type — generation, editing or extension — based on the prompt and the provided inputs.

Required. Whether the generated video includes audio synchronized with the visuals.

0 / 5
Maximum 5 video allowed
Delete All

Optional. 1-5 reference videos, each 2-30 seconds, 30 seconds in total. Use them to edit or extend a video, or to copy its motion. The provider may sometimes reject human faces. Final price is calculated from actual token usage after generation, and input video duration is charged as well.

Without a reference video: $7.00 per 1M tokens. With a reference video: $4.30 per 1M tokens. Input and output tokens are same rate.

Sample outputs
Updated Nov 16, 2026, 11:40

Overview

Seedance v2.5 Flash Reference Uncensored is a reference-guided video model by ByteDance. It turns your prompt plus visual and audio references into a new video clip, or edits an existing clip. You can steer identity, scene style, and motion by pointing the prompt at specific reference slots. It’s useful when you need repeatable creative direction across takes, without rebuilding a scene from scratch.

What you can build

  • Product ads where a specific product design stays consistent across shots
  • Character-driven clips that keep wardrobe, props, or locations from reference images
  • Motion studies where a reference video guides camera movement and pacing
  • Video edits that replace objects, change lighting, or restyle a source clip
  • Short music-video style sequences with audio that matches the on-screen action

Inputs

  • A written prompt that describes the scene, actions, camera, and audio intent. If you add references, call them out in the prompt as [Image 1], [Video 1], and [Audio 1].
  • Optional reference images (1–10) to lock identity, wardrobe, props, set design, or overall look. Use common image files like JPG, PNG, or WebP. Very large images or images with real human faces may be rejected.
  • Optional reference videos (1–5) to guide motion, pacing, blocking, or to act as the clip you want to edit or extend. Each clip should be about 2–30 seconds, and the combined total should be 30 seconds or less.
  • Optional reference audio clips (1–5) to guide timing, rhythm, ambience, or voice style. Each clip should be about 2–30 seconds, and the combined total should be 30 seconds or less.
  • A target output resolution, limited to 480p or 720p for this Flash tier.
  • A target clip length, either a fixed whole number of seconds (4–30) or Auto. Auto is the safest choice when you’re editing an existing reference clip.
  • An audio toggle to include a generated soundtrack that matches the visuals, or to return silent video.
  • An optional aspect ratio choice, or Adaptive to let the model decide. If you use a reference video, the output keeps that video’s aspect ratio.
  • An optional output container choice: MP4 for broad compatibility, or MOV for higher color precision workflows.
  • An optional watermark toggle to add an AI watermark overlay in the lower-right corner.

Outputs

  • A rendered video file in MP4 or MOV format.
  • If audio is enabled, the video includes an audio track generated in the same pass as the visuals.
  • The response typically includes a downloadable video URL plus basic file metadata (like the file name and size).

Recommended settings

  • Use 480p for quick drafts and prompt iteration. Use 720p when you want a cleaner final.
  • Pick MP4 when you need maximum playback compatibility. Pick MOV if you plan to color grade, key, or composite.
  • When you edit or extend a reference video, keep the aspect ratio adaptive and let the duration stay on Auto.

Limitations

  • Flash outputs max 30 seconds per generation.
  • Flash only supports 480p and 720p output.
  • You can attach up to 10 reference images, 5 reference videos, and 5 reference audio clips.
  • Reference video inputs must meet strict technical bounds. Out-of-range frame rates, odd aspect ratios, or huge dimensions can fail.
  • The provider may reject inputs that contain real human faces, even when the prompt is allowed.
  • Low-quality references can cause drift. Blurry frames, heavy compression, and noisy audio reduce alignment.

Safety & compliance

  • This “uncensored” listing reduces extra filtering and avoids prompt rewriting, but it is not a free-for-all.
  • Don’t request illegal sexual content, any content involving minors, or non-consensual content.
  • Don’t upload or generate identifiable real people without clear rights and consent.
  • Don’t request copyrighted characters, logos, or scenes you don’t have rights to. The model or provider may block outputs that trigger copyright or likeness checks.

Example prompts

Great starting points for seedance v2.5 flash reference uncensored.

[Image 1] the woman reclines on the red velvet chaise in the Milan loft studio and changes poses toward the camera during a boudoir photoshoot, softbox lights flashing and shutter clicks, slow dolly-in, 85mm, shallow depth of field.Video Generation
[Image 1] the woman walks slowly through the stone arches of the hacienda courtyard in her emerald lace lingerie, then stops and poses against a pillar under the pink bougainvillea in warm afternoon sun, footsteps and birdsong audio.Video Generation
[Image 1] the woman sits on the edge of the glowing pool at night, gently kicks the water, then leans back on her hands and looks into the camera, ripples of light dancing on her skin, quiet water sounds.Video Generation
[Image 1] the woman, keeping her face, hair and lavender lace lingerie from the image, wearing high nude stiletto heels, walks toward the camera down a glossy runway at a Paris lingerie fashion show in one continuous shot with no cuts. Confident catwalk stride, hips swaying, hair moving, the audience on both sides in soft darkness. She reaches the end of the runway, stops, turns her shoulder and holds a pose while photographers' cameras flash. The camera tracks backward in front of her at chest height the whole time, 50mm, warm runway spotlights, crowd murmur and camera shutter audio.Video Generation

API quick start

Run seedance v2.5 flash reference uncensored with a single API call.

POST https://api.wiro.ai/v1/Run/bytedance/seedance-v2-5-flash-reference-uncensored
{
  "prompt": "[Image 1] the woman reclines on the red v…",
  "resolution": "480p",
  "duration": "auto",
  "generateAudio": "false"
}
curl
curl -X POST "https://api.wiro.ai/v1/Run/bytedance/seedance-v2-5-flash-reference-uncensored" \
  -H "Content-Type: application/json" \
  -H "x-api-key: YOUR_WIRO_API_KEY" \
  --data-binary @- <<'JSON'
{
  "prompt": "[Image 1] the woman reclines on the red v…",
  "resolution": "480p",
  "duration": "auto",
  "generateAudio": "false"
}
JSON

Other models you might like

Explore similar video generation models.

View all
Text to Video

seedance-v1-pro-fast

bytedance

Text to Video

seedance v2.5 flash uncensored

bytedance

Text to Video

seedance v2.5 flash

bytedance

Text to Video

seedance-2.5-uncensored

bytedance

  • seedance 2.5bytedance
  • Seedance 2.0 Fastbytedance
  • Seedance 2.0bytedance
  • seedance-pro-v1bytedance
  • seedance 2.0 minibytedance
  • seedance v2.5 flash referencebytedance
  • h3 highspeedminimax
  • grok-imagine-videoxai

Discover, test, and run AI models, build workflows and agents with one unified API.

All systems operational
WiroModelsAgentsMCP ServerLLM GatewayBrowser ExtensionIntegrationsPricingAboutBlogCareersContactPartnerChangelogStatusFAQ
CategoriesText to ImageImage to ImageImage EditingImage to TextText to VideoImage to VideoVideo to VideoText to SpeechSpeech to TextSpeech to SpeechVoice CloneRealtime ConversationMusic GenerationLLMUtilities
ModelsNano Banana 2GPT Image 2.5Seedream V5 ProSeedance 2.5Veo 3.1Kling V3FLUX 3FLUX.2 ProWan 3.0 PrimeGrok Imagine 1.5
PartnersGoogleOpenAIByteDanceBlack Forest LabsKling AIQwenAlibabaxAIMiniMaxElevenLabs
Getting StartedWiro ModelsWiro EffectsIntroductionAuthenticationProjectsCode ExamplesWiro MCP ServerSelf-Hosted MCPn8n IntegrationLLMs.txt
API ReferenceModelsRun a ModelModel ParametersTasksLLM & Chat StreamingWebSocketRealtime VoiceFiles
© 2026 Wiro AI. All rights reserved.
PrivacyTermsData Deletion