Asynchronous AI processing pipeline designed for high-performance automation.
This workflow has no configurable inputs. Check the visual builder to add task inputs.
About this workflow
This workflow is an asynchronous AI processing pipeline. Submit your inputs and receive results when the pipeline completes.
Use via REST API
Submit inputs to POST /workflow/street-portrait-photography-seedream-v5-chroma/api-execute and poll GET /workflow/run/{run_id}/status until the run completes. Switch to the for full code examples.
This workflow generates high-quality editorial street portrait photographs — the fashion-forward, location-shot style seen in editorial magazines and social fashion content. LLM reasoning scripts the scene and styling concept, Seedream v5 generates the initial fashion portrait, Chroma applies colour grading and cinematic colour treatment, Nano Banana 2 handles compositing and background refinement, and Qwen Image applies final retouching.
The multi-model colour pipeline (Seedream v5 → Chroma → Qwen) produces the distinctive editorial colour signature that distinguishes professional street photography from AI-generated content.
Scene concept
LLM reasoning develops the styling, location type, and editorial direction.
Portrait generation
Seedream v5 renders the initial street portrait with the styled model and location context.
Colour grading
Chroma applies cinematic colour treatment for the editorial film look.
Compositing and retouching
Nano Banana 2 refines the scene composite and Qwen Image applies final portrait retouching.