Combine My References

Choose a starting point

Pick the result you want. Quick mode handles the production choices, while Standard and Pro remain available in the builder.

Browse all

Real preview

Dynamic Visual Seed

Create the first reusable visual asset for a content system or guide.

6 credits · image

Real preview

Turn Source Into Visuals

Generate source-faithful images from a passage, script, lesson, sermon note, lore brief, or episode topic.

8 credits · image set

Real preview

Build A Character Sheet

Generate repeatable character reference images for comics, stories, series, and Seedance source frames.

8 credits · image set

Real preview

Make Story Panels

Generate story panel images from a source, script, scene, or reference stack.

8 credits · image set

Real preview

Visual Storyboard

Turn a scene, script, lesson, or campaign idea into ordered storyboard images.

8 credits · image set

Real preview

Combine My References

Combine face, outfit, pose, setting, and style references into one reusable keyframe.

6 credits · image

Combine My References

Combine face, outfit, pose, setting, and style references into one reusable keyframe.

6 credits

Add the essentials and let Promptha choose the production settings.

Starting points

Scene briefrequired

Style

Details

References

Use accepted outputs or suggested Creative DB references. Raw query syntax stays hidden.

0 selected

Face

identity, face, expression

media

Upload

Outfit

wardrobe, fabric, silhouette

media

Upload

Pose

gesture, stance, action start

media

Upload

Setting

location, background, lighting

media

Upload

Generation settings

FormatComposition

Scene brief is required before this block can run.

Output

characterSceneImage

imageWorkspace

No output yet

The generated block output will appear here after the block runs.

Recent runs

Workspace

Loading recent block runs...

Block framework

This is the production contract for the block: prompt, settings, model route, output shape, QA gates, storage, and guide compatibility.

Production ready

Passed

Definition

Passed

Minimal Ux

Passed

Standalone Resolve

Passed

Guide Smoke

Passed

Media Artifacts

Warnings

Non-reference-first blocks should not surface multiple media reference slots in the default run layer.

Input contract

Scene brief

sceneBrief · markdown · required

Describe the subject, action, place, emotional point, and what the final frame should communicate.

Output contract

Character scene image

characterSceneImage · image · required

One generated scene image that can be accepted as a video source frame.

Visible settings

Format

aspectRatio · visible

The output frame shape. Default: 16:9.

Composition

composition · visible

Primary frame language. Default: cinematic-wide.

Reference strength

referenceStrength · advanced

Advanced: how strongly references should constrain the image. Default: balanced.

Avoid

negativeRules · advanced

Advanced: artifacts or content to avoid. Default: No fake text, no logos, no watermark, no distorted anatomy, no duplicate subjects, no unreadable signs, no cluttered background..

Model and services

Default model

@promptha.recraft-v4-pro

Reference-aware single image keyframes for character, local business, product, and story scenes.

askgl.image.generate

ask · promptha · required

Routes through Promptha image generation from the resolved multi-reference AskGL prompt.

reference.stack.resolve

creative-db · promptha

Creative DB visual style references are selected visually and injected into the resolved prompt.

Quality gates

Single image output

single-image-output · blocking

The block must return one image URL for characterSceneImage.

Reference coherence

reference-coherence · blocking

References should merge into one coherent scene rather than a collage.

Video source ready

video-source-ready · blocking

The output should be stable and readable enough to feed a video block.

Video source provenance

video-source-provenance · blocking

Any character keyframe sent to video must remain truthfully identified as an original AI-generated fictional character or an owned/authorized human reference.

Compatibility

Feeds into

reference-to-video-generator

sourceImageUrl -> sourceImageUrl

The generated character scene image can become the primary image for reference-to-video.

Feeds into

seedance-scene-video-director

sourceImageUrl -> sourceImageUrl

The generated keyframe can feed the practical Seedance scene video block.

Storage and examples

Storage

characterSceneImage · run-state, ask-output, saved-output

Run state key: nodes.characterSceneImage. Saved output kind: image.

Local founder keyframe

local-founder-keyframe

One coherent cinematic keyframe that can feed a Seedance reference-to-video block.

Blocks

Combine My References

Choose a starting point

Pick the result you want. Quick mode handles the production choices, while Standard and Pro remain available in the builder.

Browse all

Real preview

Dynamic Visual Seed

Create the first reusable visual asset for a content system or guide.

6 credits · image

Real preview

Turn Source Into Visuals

Generate source-faithful images from a passage, script, lesson, sermon note, lore brief, or episode topic.

8 credits · image set

Real preview

Build A Character Sheet

Generate repeatable character reference images for comics, stories, series, and Seedance source frames.

8 credits · image set

Real preview

Make Story Panels

Generate story panel images from a source, script, scene, or reference stack.

8 credits · image set

Real preview

Visual Storyboard

Turn a scene, script, lesson, or campaign idea into ordered storyboard images.

8 credits · image set

Real preview

Combine My References

Combine face, outfit, pose, setting, and style references into one reusable keyframe.

6 credits · image

Combine My References

Combine face, outfit, pose, setting, and style references into one reusable keyframe.

6 credits

Add the essentials and let Promptha choose the production settings.

Starting points

Scene briefrequired

Style

Details

References

Use accepted outputs or suggested Creative DB references. Raw query syntax stays hidden.

0 selected

Face

identity, face, expression

media

Upload

Outfit

wardrobe, fabric, silhouette

media

Upload

Pose

gesture, stance, action start

media

Upload

Setting

location, background, lighting

media

Upload

Generation settings

FormatComposition

Scene brief is required before this block can run.

Output

characterSceneImage

imageWorkspace

No output yet

The generated block output will appear here after the block runs.

Recent runs

Workspace

Loading recent block runs...

Block framework

This is the production contract for the block: prompt, settings, model route, output shape, QA gates, storage, and guide compatibility.

Production ready

Passed

Definition

Passed

Minimal Ux

Passed

Standalone Resolve

Passed

Guide Smoke

Passed

Media Artifacts

Warnings

Non-reference-first blocks should not surface multiple media reference slots in the default run layer.

Input contract

Scene brief

sceneBrief · markdown · required

Describe the subject, action, place, emotional point, and what the final frame should communicate.

Output contract

Character scene image

characterSceneImage · image · required

One generated scene image that can be accepted as a video source frame.

Visible settings

Format

aspectRatio · visible

The output frame shape. Default: 16:9.

Composition

composition · visible

Primary frame language. Default: cinematic-wide.

Reference strength

referenceStrength · advanced

Advanced: how strongly references should constrain the image. Default: balanced.

Avoid

negativeRules · advanced

Advanced: artifacts or content to avoid. Default: No fake text, no logos, no watermark, no distorted anatomy, no duplicate subjects, no unreadable signs, no cluttered background..

Model and services

Default model

@promptha.recraft-v4-pro

Reference-aware single image keyframes for character, local business, product, and story scenes.

askgl.image.generate

ask · promptha · required

Routes through Promptha image generation from the resolved multi-reference AskGL prompt.

reference.stack.resolve

creative-db · promptha

Creative DB visual style references are selected visually and injected into the resolved prompt.

Quality gates

Single image output

single-image-output · blocking

The block must return one image URL for characterSceneImage.

Reference coherence

reference-coherence · blocking

References should merge into one coherent scene rather than a collage.

Video source ready

video-source-ready · blocking

The output should be stable and readable enough to feed a video block.

Video source provenance

video-source-provenance · blocking

Any character keyframe sent to video must remain truthfully identified as an original AI-generated fictional character or an owned/authorized human reference.

Compatibility

Feeds into

reference-to-video-generator

sourceImageUrl -> sourceImageUrl

The generated character scene image can become the primary image for reference-to-video.

Feeds into

seedance-scene-video-director

sourceImageUrl -> sourceImageUrl

The generated keyframe can feed the practical Seedance scene video block.

Storage and examples

Storage

characterSceneImage · run-state, ask-output, saved-output

Run state key: nodes.characterSceneImage. Saved output kind: image.

Local founder keyframe

local-founder-keyframe

One coherent cinematic keyframe that can feed a Seedance reference-to-video block.