HiNow LogoHiNow

Lucy 2.5

Model Information

Display Name: Lucy 2.5

API Model ID: decart/lucy-2.5

Category: Video To Video

Description: Lucy 2.5 is Decart's flagship video editing model. Upload an existing video, describe the change (or provide a reference image, or both) and Lucy rewrites the footage while preserving the original motion, temporal consistency and scene structure — clothing changes, character/object replacement, background swaps and style transfers. **Key Features:** - Video-to-video editing driven by text, a reference image, or both - Preserves original motion patterns and temporal consistency - Unlimited input duration (no clip limit) - Self-anchoring on the model's own latent for stable results (on by default) - Optional automatic prompt enhancement **Capabilities:** - Character and object replacement - Clothing and accessory changes - Background replacement - Style transfer across the whole clip **Best For:** - Editing existing footage instead of regenerating it - Product/character swaps in marketing videos - Consistent restyling of long clips **Technical Specs:** - Input: MP4 (H.264 or VP8), 16:9 or 9:16, up to 200 MB - Output: 720p MP4 (1280×720 or 720×1280), same duration as input - Prompt: best results with 20-40 words focused on one specific change **Parameters:** | Parameter | Type | Required | Description | |-----------|------|----------|-------------| | video | URL / base64 | Yes | Input video to edit (in `parameters.video`) | | prompt | string | No* | Edit instruction (*required unless a reference image is given) | | images[0] / reference_image | URL / base64 | No | Reference image guiding the edit | | seed | integer | No | 0..4294967295 | | enhance_prompt | boolean | No | Auto-enhance the prompt (default true) | | self_anchor | boolean | No | Re-anchor on the model's own latent (default true) | **Pricing:** billed per second of input video ($0.06/s).

How to Use This Model

To use Lucy 2.5 via the HInow.ai API, use the model ID: decart/lucy-2.5

API Request Example


POST https://api.hinow.ai/v1/chat/completions
Authorization: Bearer YOUR_API_KEY
Content-Type: application/json

{
  "model": "decart/lucy-2.5",
  "messages": [
    {"role": "user", "content": "Your input here"}
  ]
}

Pricing

  • total: $0.06

Available Parameters

  • video: Input video to edit (URL or base64 data URI). MP4 H.264/VP8, 16:9 or 9:16, max 200 MB. Billed per second of this video.
  • reference_image: Optional reference image (URL or base64) guiding the edit. images[0] is accepted as an alias.
  • resolution: Output resolution (only 720p). (Options: 720p)
  • seed: Random seed for reproducibility (0-4294967295).
  • enhance_prompt: Let Decart rewrite/enhance the prompt automatically.
  • self_anchor: Re-anchor on the model's own latent for stability (Lucy 2.5 self-anchor).

Quick Reference

To use this model, set: "model": "decart/lucy-2.5"

Featured: No

Documentation: https://hinow.ai/models/decart/lucy-2.5

API Endpoint: https://api.hinow.ai/v1

Back to Models

Lucy 2.5

decart/lucy-2.5

$0.060
per second

About

Lucy 2.5 is Decart's flagship video editing model. Upload an existing video, describe the change (or provide a reference image, or both) and Lucy rewrites the footage while preserving the original motion, temporal consistency and scene structure — clothing changes, character/object replacement, background swaps and style transfers.

Key Features:

  • Video-to-video editing driven by text, a reference image, or both
  • Preserves original motion patterns and temporal consistency
  • Unlimited input duration (no clip limit)
  • Self-anchoring on the model's own latent for stable results (on by default)
  • Optional automatic prompt enhancement

Capabilities:

  • Character and object replacement
  • Clothing and accessory changes
  • Background replacement
  • Style transfer across the whole clip

Best For:

  • Editing existing footage instead of regenerating it
  • Product/character swaps in marketing videos
  • Consistent restyling of long clips

Technical Specs:

  • Input: MP4 (H.264 or VP8), 16:9 or 9:16, up to 200 MB
  • Output: 720p MP4 (1280×720 or 720×1280), same duration as input
  • Prompt: best results with 20-40 words focused on one specific change

Parameters:

ParameterTypeRequiredDescription
videoURL / base64YesInput video to edit (in parameters.video)
promptstringNo*Edit instruction (*required unless a reference image is given)
images[0] / reference_imageURL / base64NoReference image guiding the edit
seedintegerNo0..4294967295
enhance_promptbooleanNoAuto-enhance the prompt (default true)
self_anchorbooleanNoRe-anchor on the model's own latent (default true)

Pricing: billed per second of input video ($0.06/s).

Capabilities

Video To Video

Parameters

video

Input video to edit (URL or base64 data URI). MP4 H.264/VP8, 16:9 or 9:16, max 200 MB. Billed per second of this video.

reference_image

Optional reference image (URL or base64) guiding the edit. images[0] is accepted as an alias.

resolution

Output resolution (only 720p).

720p
seed

Random seed for reproducibility (0-4294967295).

enhance_prompt

Let Decart rewrite/enhance the prompt automatically.

self_anchor

Re-anchor on the model's own latent for stability (Lucy 2.5 self-anchor).

Code Examples

curl -X POST https://api.hinow.ai/v1/videos \
  -H "Content-Type: application/json" \
  -H "Authorization: Bearer $HINOW_API_KEY" \
  -d '{
    "model": "decart/lucy-2.5",
    "prompt": "Enhance and upscale this video",
    "video_url": "https://example.com/video.mp4",
    "parameters": {
      "video": "",
      "reference_image": "",
      "resolution": "720p",
      "seed": "",
      "enhance_prompt": "",
      "self_anchor": ""
    }
  }'